Vapor has a direction, and the direction changes with the climate.
Steep-slope and low-slope roofs · cooling-dominated and mixed-humid climates
Almost every published roof-moisture explanation is written for a cold climate. Run the same detail in a hot-humid one and the vapor retarder ends up on the wrong side of the assembly.
Why does a roof assembly that works in a cold climate grow mold in a humid one?
Vapor moves from warm and humid toward cool and dry, so its direction reverses with climate and season. In a cooling-dominated hot-humid climate, humid outdoor air meets air-conditioned surfaces and condenses inside the assembly — the opposite of the cold-climate case most guidance describes. A vapor retarder detailed for a cold climate ends up on the wrong side, and mold follows the moisture, not the roof.
The short versionSection link
The code language quoted on this page is model text — provisions a state or local government may adopt, amend, delay, or decline — described in a U.S. Department of Energy code-compliance brief. It is not the law where you live, and this site did not read the International Code Council’s own published text directly. Confirm the adopted edition, its amendments, and its effective date with your authority having jurisdiction before treating anything here as a requirement.
- The rule that governs everything
- Any surface colder than the dew point of the air touching it will collect waterThe National Weather Service defines dew point as the temperature air must be cooled to, at constant pressure, to reach 100% relative humidity. That single definition is the whole mechanism.
- Which way vapor drives in a hot-humid cooling season
- Inward — from outdoors toward the conditioned spaceDOE's Building America Solution Center: “In hot, humid locations, the outside air is likely to be more humid than the inside air. Inside surfaces may be cooler and may provide condensing surfaces for humid air introduced into a vented attic, thus encouraging condensation in the attic.”
- The condensing surface in a heating climate
- The roof sheathingBASC: “The key to creating an unvented roof assembly is to keep the roof deck – the principle condensing surface in roof assemblies – sufficiently warm throughout the year such that condensation will not occur.”
- The condensing surface in a cooling climate
- The air-conditioning equipment and the chilled ceiling planeBuilding Science Corporation: “In hot humid climates condensation on ductwork and air handlers located in vented attics is common.” The deck is the hot side, not the cold one.
- Interior vapor retarders in a warm-humid climate
- The default advice is not to install oneBASC: “In warm-humid climates, do not install Class 1 vapor retarders on the interior side of air-permeable insulation in above-grade walls, except at shower and tub walls.” Joseph Lstiburek puts it more broadly: “If you air-condition and you have humid summers there should be no interior vapor barrier.”
- What the model code conditions on
- Climate zone — and it asks for opposite things at the two endsIn DOE's summary of the model unvented-attic provisions, climate zones 5 to 8 require air-impermeable insulation to be a Class II vapor retarder or to carry one, while climate zones 1, 2 and 3 are offered a vapor-open ridge detail instead. Model text; confirm the adopted edition locally.
- What hot-humid means, technically
- >20 in annual precipitation, plus a wet-bulb thresholdDOE's Building America definition: a 67°F or higher wet-bulb temperature for 3,000+ hours, or 73°F or higher for 1,500+ hours, during the warmest six consecutive months.
- Why remediation alone fails
- The growth returns if the water does not stopEPA: “The key to mold control is moisture control.” Clean the mold without fixing the water problem and, in the EPA's words, most likely the mold problem will come back.
- Indoor humidity target
- Below 60% RH, ideally 30–50%EPA's stated range for mold prevention. It is an indoor-air target, not a roof specification, and it does not by itself tell you whether a specific assembly is safe.
| Situation | Dominant vapor drive | Coldest surface — where condensation lands | What the model provisions key on |
|---|---|---|---|
| Cold climate, heating season | Inside → out. Heated, humidified indoor air is more humid than the cold outdoor air. | The roof sheathing, which is close to outdoor temperature. | Keeping the sheathing warm enough. In DOE's summary, zones 5–8 require air-impermeable insulation to be a Class II vapor retarder or to have one in direct contact with its underside. A separate provision, not limited to those zones, governs assemblies using air-permeable insulation below the deck: rigid board or sheet insulation above the sheathing, either at the R-value the model table sets for the climate zone or, alternatively, enough of it to hold the monthly average temperature of the sheathing underside above 45°F. |
| Hot-humid climate, cooling season | Outside → in. Outdoor air is more humid than the air-conditioned interior. | AC ducts, the air handler, and the chilled top face of the ceiling plane. The deck is the hot side. | Letting the assembly dry, rather than sealing it. For zones 1, 2 and 3 with air-permeable insulation, DOE's brief describes a vapor diffusion port near the ridge, ≥20 perms tested to ASTM E96 Procedure A, at not more than 12 inches from the highest point of the roof and an area ≥1:600 of the ceiling area. |
| Mixed-humid climate | Both, in different seasons. The direction reverses twice a year. | It moves. The sheathing in winter; the conditioned surfaces in summer. | Nothing on this page resolves it. An assembly that has to dry in both directions is a design problem for a professional, not a table lookup. |
| Vented attic, ducts in the attic | Outside → in, straight through the vents. | The ductwork itself, which is below the dew point of the air around it whenever the system runs. | Not a code question so much as a location question. BASC: “If HVAC equipment is located in the attic, an unvented assembly will be much more efficient because the ductwork and air handler will be inside the thermal envelope.” |
| Unvented, conditioned attic | Controlled. There is no ventilation air, so the moisture that matters is what leaks or diffuses in. | Whatever the design left cold. In an unvented roof the deck itself becomes the surface to keep warm — or, in hot-humid zones, the surface to let dry. | The full set of model conditions in DOE's brief: the space is completely within the building thermal envelope, no interior Class I vapor retarder is installed on the ceiling side of the assembly, insulation type and placement, and the climate-zone-specific vapor and diffusion requirements above. |
Read this table one item at a time
Cold climate, heating season
- Dominant vapor drive
- Inside → out. Heated, humidified indoor air is more humid than the cold outdoor air.
- Coldest surface — where condensation lands
- The roof sheathing, which is close to outdoor temperature.
- What the model provisions key on
- Keeping the sheathing warm enough. In DOE's summary, zones 5–8 require air-impermeable insulation to be a Class II vapor retarder or to have one in direct contact with its underside. A separate provision, not limited to those zones, governs assemblies using air-permeable insulation below the deck: rigid board or sheet insulation above the sheathing, either at the R-value the model table sets for the climate zone or, alternatively, enough of it to hold the monthly average temperature of the sheathing underside above 45°F.
Hot-humid climate, cooling season
- Dominant vapor drive
- Outside → in. Outdoor air is more humid than the air-conditioned interior.
- Coldest surface — where condensation lands
- AC ducts, the air handler, and the chilled top face of the ceiling plane. The deck is the hot side.
- What the model provisions key on
- Letting the assembly dry, rather than sealing it. For zones 1, 2 and 3 with air-permeable insulation, DOE's brief describes a vapor diffusion port near the ridge, ≥20 perms tested to ASTM E96 Procedure A, at not more than 12 inches from the highest point of the roof and an area ≥1:600 of the ceiling area.
Mixed-humid climate
- Dominant vapor drive
- Both, in different seasons. The direction reverses twice a year.
- Coldest surface — where condensation lands
- It moves. The sheathing in winter; the conditioned surfaces in summer.
- What the model provisions key on
- Nothing on this page resolves it. An assembly that has to dry in both directions is a design problem for a professional, not a table lookup.
Vented attic, ducts in the attic
- Dominant vapor drive
- Outside → in, straight through the vents.
- Coldest surface — where condensation lands
- The ductwork itself, which is below the dew point of the air around it whenever the system runs.
- What the model provisions key on
- Not a code question so much as a location question. BASC: “If HVAC equipment is located in the attic, an unvented assembly will be much more efficient because the ductwork and air handler will be inside the thermal envelope.”
Unvented, conditioned attic
- Dominant vapor drive
- Controlled. There is no ventilation air, so the moisture that matters is what leaks or diffuses in.
- Coldest surface — where condensation lands
- Whatever the design left cold. In an unvented roof the deck itself becomes the surface to keep warm — or, in hot-humid zones, the surface to let dry.
- What the model provisions key on
- The full set of model conditions in DOE's brief: the space is completely within the building thermal envelope, no interior Class I vapor retarder is installed on the ceiling side of the assembly, insulation type and placement, and the climate-zone-specific vapor and diffusion requirements above.
Model code text has no jurisdiction and no effective date of its own. It becomes enforceable only where a government adopts it, on that government's effective date, and adoptions routinely amend it. This site read DOE's code-compliance brief, not the ICC's published code.
This page's position — and the situations where following it would be a mistakeSection link
The position here is that in a cooling-dominated climate you should stop trying to keep moisture out of an assembly with a barrier and start asking which surface is cold and whether the assembly can dry. That framing is right often enough to be the default. It is not right always.
Best when
- The air handler, the ductwork, or both are in the attic. That is a cold-surface problem sitting inside a hot, humid space, and no amount of ventilation air fixes it.
- The building is in a genuinely cooling-dominated climate, where the humid season is long and the heating season is short enough that the outward drive barely runs.
- Somebody has already installed a polyethylene sheet or a foil-faced product on the room side of the ceiling in a warm-humid climate, and the assembly now has no direction in which it can dry.
- The roof is coming off anyway, so the vapor strategy, the insulation, the air barrier, and the underlayment can all be decided together instead of one at a time.
- Bathroom or kitchen exhaust fans discharge into the attic rather than outdoors, which is a moisture source added deliberately, on purpose, every day.
- The mold has been cleaned once already and came back. That is the diagnostic fact that matters most on this whole page.
Think twice if
- You are in a mixed-humid climate. The drive reverses seasonally, an assembly has to dry in two directions, and the hot-humid answer applied year-round is as wrong as the cold-climate answer applied year-round.
- The attic is vented, the ceiling plane is reachable and reasonably tight, and there is no mechanical equipment up there. Converting a working vented attic into an unvented one is a large expense to solve a problem you may not have.
- You want to be able to find a roof leak. BASC is direct about the tradeoff: “A vented attic insulated at the attic floor makes it easier to detect and solve roof leaks or other moisture-related issues.” Spray foam against the deck removes that.
- The plan is to spray foam under an old deck. You are gluing an insulation layer to sheathing you can no longer inspect, and you are complicating the next tear-off for whoever does it.
- Someone is selling a vapor diffusion port as the whole answer. The DOE Building America report on this exact detail says diffusion ports “should not be widely recommended as the sole method to mitigate attic moisture issues” pending further investigation.
- The house predates 1990 and the work would disturb old felts, mastics, or pipe and duct insulation. Those may contain asbestos, and testing precedes disturbance.
- The moisture is arriving as bulk water. A flashing failure or a leaking penetration is a roofing repair, and no vapor strategy on earth fixes a hole.
- You are renting, or selling this year. Rebuilding the moisture control of a roof assembly is a decade-scale decision with a nine-week payback of zero.
What changes the answer
- Climate zone and its moisture designation — the letter after the number is doing at least as much work as the number.
- Where the mechanical equipment lives. Ducts in an unconditioned attic change the answer more than the roof covering does.
- The thermostat setpoint and how the household actually runs the house. The modelling reported alongside the DOE field study of this assembly showed “high sensitivity to occupancy conditions.”
- Whether the assembly is vented or a correctly designed unvented one. Both are legitimate; they fail differently and are fixed differently.
- Whether an interior vapor retarder already exists, and what class it is.
- Indoor moisture generation: unvented bathroom fans, unsealed crawl spaces, aquariums, indoor drying, and a dehumidifier that has not run since 2019.
- Whether the deck can dry outward at all, which depends on the underlayment and the covering above it as well as on what is below.
- The adopted code edition and local amendments, which decide which of the model provisions above is even in play at your address.
The same detail, run in two directionsSection link
Water vapor does not know which climate it is in. It moves from more humid toward less humid, and it turns to liquid on whatever surface is below its dew point. Everything else on this page follows from those two sentences.
The National Weather Service defines dew point as “the temperature the air needs to be cooled to (at constant pressure) in order to achieve a relative humidity (RH) of 100%.” Read that as an instruction rather than a definition and it becomes useful: name the dew point of the air, then find every surface colder than it. Those surfaces are where the water will be.
The NWS also gives a rough comfort scale for summer dew points: at or below 55 °F is dry and comfortable, 55 to 65 is becoming sticky, and 65 °F and above means “lots of moisture in the air, becoming oppressive.” A Gulf or South Atlantic afternoon spends much of its summer in that top band. Air at a 75 °F dew point will wet anything colder than 75 °F that it touches, and in an air-conditioned house there is a great deal of that.
1 · In a heating climate, the house is the humid side
This is the case nearly every roofing article describes, because nearly every roofing article was written for it. Indoor air in winter is warm and carries moisture from cooking, washing, breathing, and plants. Outdoor air is cold and dry. Vapor pushes outward, air leakage carries far more of it than diffusion does, and the first cold thing it meets is the underside of the roof deck.
DOE’s Building America Solution Center states the design objective for that case plainly: “The key to creating an unvented roof assembly is to keep the roof deck – the principle condensing surface in roof assemblies – sufficiently warm throughout the year such that condensation will not occur.” Everything in cold-climate roof detailing follows from that sentence — the interior vapor retarder, the rigid insulation above the deck, the ventilation that flushes the space. The cold surface is the deck, so you either warm it or you keep vapor away from it.
2 · In a cooling climate, the outdoors is the humid side
Run an air conditioner in a hot, humid place and the gradient reverses. BASC states the condition directly: “In hot, humid locations, the outside air is likely to be more humid than the inside air. Inside surfaces may be cooler and may provide condensing surfaces for humid air introduced into a vented attic, thus encouraging condensation in the attic.”
Notice what that sentence does to the roof. The deck is not the cold surface any more; in the sun it is the hottest thing on the property. The cold surfaces are inside the building — the chilled ceiling plane, the supply ducts, the air handler cabinet. Building Science Corporation’s Joseph Lstiburek reports the consequence as ordinary rather than exotic: “In hot humid climates condensation on ductwork and air handlers located in vented attics is common.”
3 · Which is why the vapor retarder ends up on the wrong side
A vapor retarder works by being placed on the side of the assembly that is warm during the season when vapor drives hardest. In a cold climate that is the interior. Install the same layer in a cooling-dominated climate and it sits on the cold side of the summer gradient: vapor arriving from outdoors reaches it, stops, and has nowhere to go, because the layer that was supposed to keep moisture out is now the thing keeping moisture in.
BASC’s guidance for that climate is a prohibition rather than a preference: “In warm-humid climates, do not install Class 1 vapor retarders on the interior side of air-permeable insulation in above-grade walls, except at shower and tub walls.” Lstiburek generalises it past walls and past the deep South — “If you air-condition and you have humid summers there should be no interior vapor barrier” — and notes that the layer is often installed by accident rather than by specification. Vinyl wallcoverings, he writes, “are vapor barriers. Really. They are made out of vinyl,” with a Type II covering measuring “less than 1 perm.”
A perm is the unit of vapor permeance. Class I is 0.1 perm or less — effectively closed. What matters is not the number in isolation but which face of the assembly it is on relative to the direction the vapor is travelling, and that is a climate question.
4 · Ventilation is not a universal answer here, in either direction
In a cold climate, ventilation air is dry, so flushing the attic with it removes moisture. In a hot-humid climate, the ventilation air is the moisture. That is the reason the site’s standing rule — that there is no universal ventilation ratio and that “more ventilation is always better” is false — lands hardest on this page. BASC lists the factors that decide it: “climate,” “desire for additional living and storage space,” “building design and configuration,” and “location of HVAC.” None of those is a number you can look up.
The arithmetic of net free area, the balance between intake and exhaust, and the mixed-exhaust mistakes belong on the ventilation guide, which works them through on a real house and states where the unvented alternative changes the rules. This page is about which direction the moisture is coming from, not how big the vents are.
5 · So an unvented assembly is a legitimate answer here
A correctly designed unvented attic moves the insulation, air control, and thermal boundary to the roof plane, so the attic becomes part of the conditioned space. BASC gives two hot-climate reasons for it. First, moisture: “In hot, humid locations, it may also be advisable to construct unvented rather than vented attics to help reduce the entry of moisture-laden air into the attic.” Second, equipment: “If HVAC equipment is located in the attic, an unvented assembly will be much more efficient because the ductwork and air handler will be inside the thermal envelope.” Bring the ducts inside the envelope and they stop being cold surfaces in humid air.
That is not the same as saying unvented is better. BASC names the cost of it in the same document: “A vented attic insulated at the attic floor makes it easier to detect and solve roof leaks or other moisture-related issues,” and “if a leak does occur, moisture damage is less likely because the exposed and ventilated rafters and decking can dry out more readily.” You are trading leak visibility and drying capacity for control of the humid air. Whether that trade is worth making is a building question, not a general one.
6 · The model code splits on climate, in opposite directions
DOE publishes a code-compliance brief on unvented attics that sets out the model provisions. Two of them are worth putting side by side, because together they demonstrate that the model code itself does not believe in a universal moisture strategy.
For the cold end: “In climate zones 5, 6, 7, and 8, any air-impermeable insulation shall be a Class II vapor retarder, or shall have a Class II vapor retarder coating or covering in direct contact with the underside of the insulation.” Seal the warm side, keep the deck warm.
For the hot end: in climate zones 1, 2 and 3, an assembly using air-permeable insulation may instead use a vapor diffusion port — a deliberately vapor-open opening near the ridge, so that moisture leaves by diffusion without ventilating the space with humid outdoor air. In DOE’s wording, “An approved vapor diffusion port shall be installed not more than 12 inches (305 mm) from the highest point of the roof,” with an area ≥1:600 of the ceiling area, and “The vapor permeable membrane in the vapor diffusion port shall have a vapor permeance rating of ≥20 perms when tested in accordance with Procedure A of ASTM E96.”
One end of the country is told to close the assembly. The other is told to open a hole in the ridge on purpose. That is the whole argument of this page, written into the model code.
Two honest caveats. DOE’s brief is written against the 2015 IRC and attributes the diffusion-port language to 2018 ICC code proposals, so the edition and section numbering that apply where you live may differ — this site read DOE’s brief and did not read the ICC’s own published text. And jurisdictions edit exactly this provision: Southface Institute, a building-science nonprofit in Atlanta, reports that Georgia deleted the diffusion-port language when it adopted the 2018 IRC in 2020 and reinstated it at a later code review. This site could not confirm that account against Georgia’s own published amendment document, so treat it as what it is — an illustration that adopted text and model text are different documents, not a statement of Georgia law.
7 · The mold is downstream of all of this
Mold on roof sheathing is not a roofing defect. It is the visible record of a surface that stayed wet long enough, and the EPA’s guidance is unambiguous about what that means for the repair: “The key to mold control is moisture control.” Clean the growth and leave the water problem in place and, in the EPA’s words, most likely the mold problem will come back. It also warns against the shortcut: “Do not paint or caulk moldy surfaces. Clean up the mold and dry the surfaces before painting. Paint applied over moldy surfaces is likely to peel.”
Which is why a remediation quote that contains no sentence about where the water came from is not a quote for a solution. It is a quote for doing this again.
Find the cold surface: the only calculation on this pageSection link
The whole diagnosis reduces to one comparison, repeated once per surface. Name the dew point of the air. Name the temperature of the surface. If the surface is colder, that is where the water is.
Take a cooling-season afternoon in a humid region and assume an outdoor dew point of 75 °F. That is an assumption for the arithmetic, not a measurement of anywhere: the National Weather Service’s summer scale puts anything at or above 65 °F in its “lots of moisture in the air, becoming oppressive” band, and a Gulf or South Atlantic afternoon sits well inside it. Assume also that the thermostat is set to 74 °F and that the attic is vented, so the air in the attic is essentially outdoor air.
From the NWS definition — the dew point is the temperature air must be cooled to in order to reach 100% relative humidity — the rule falls out with no further physics: every surface that air touches which is colder than 75 °F will collect liquid water. Now go round the assembly.
| Surface | Roughly where its temperature comes from | Below the 75°F dew point? | What follows |
|---|---|---|---|
| Roof sheathing, sunlit afternoon | Solar gain on a dark covering pushes it far above outdoor air temperature. | No — it is the hottest surface in the building. | This is the surface all cold-climate guidance is about, and in this condition it is the one surface that cannot be the problem. Detailing the roof as though it were is how the wrong answer gets built. |
| Roof sheathing, clear night | Radiant loss to a clear sky, the same effect that puts dew on a car roof and not on the pavement beside it, can drop a surface below the air temperature around it. | Possibly, for part of the night. | It is why sheathing in a humid climate can be damp at dawn and dry by ten. Intermittent wetting that dries each day is a different situation from wetting that never dries, and the distinction is the whole risk assessment. |
| Air-conditioning supply duct in the attic | It carries air below room temperature by design, and duct insulation only slows the surface toward the surrounding air. | Yes, whenever the system runs. | Building Science Corporation records condensation on ductwork and air handlers in vented attics as common in hot humid climates. This is usually the first thing to find and the cheapest thing to fix. |
| Air handler cabinet in the attic | Same reason, with more surface area and more places for the insulation to be interrupted at a seam or a service panel. | Yes. | Water dripping off an air handler is frequently reported as a roof leak. It is not one, and no roofing work will change it. |
| Top face of the ceiling gypsum, under the insulation | It sits close to the temperature of the air-conditioned room below it, at the thermostat setpoint. | Yes — 74°F is below 75°F, and the margin is one degree. | Any humid attic air that reaches it through a gap in the ceiling plane will wet it. It also shows why the thermostat setpoint is a moisture variable: the modelling reported alongside the DOE field study of this assembly showed high sensitivity to occupancy conditions. |
| A Class I vapor retarder on the room side of the ceiling | Room temperature, and effectively closed to vapor at 0.1 perm or less. | Yes, and worse than yes. | It is a cold, impermeable face in the summer gradient. Vapor arriving from above stops there and cannot dry inward. BASC's warm-humid guidance is not to install one on the interior side of air-permeable insulation in the first place. |
| Interior gypsum behind a vinyl wallcovering | Room temperature, with a low-permeance finish on the room side. | Yes. | Lstiburek's summary is blunt: vinyl wall coverings “are vapor barriers. Really. They are made out of vinyl,” with a Type II covering under 1 perm. It is a wall case rather than a roof case, and it is the clearest illustration on this page of a vapor retarder installed by accident. |
Read this table one item at a time
Roof sheathing, sunlit afternoon
- Roughly where its temperature comes from
- Solar gain on a dark covering pushes it far above outdoor air temperature.
- Below the 75°F dew point?
- No — it is the hottest surface in the building.
- What follows
- This is the surface all cold-climate guidance is about, and in this condition it is the one surface that cannot be the problem. Detailing the roof as though it were is how the wrong answer gets built.
Roof sheathing, clear night
- Roughly where its temperature comes from
- Radiant loss to a clear sky, the same effect that puts dew on a car roof and not on the pavement beside it, can drop a surface below the air temperature around it.
- Below the 75°F dew point?
- Possibly, for part of the night.
- What follows
- It is why sheathing in a humid climate can be damp at dawn and dry by ten. Intermittent wetting that dries each day is a different situation from wetting that never dries, and the distinction is the whole risk assessment.
Air-conditioning supply duct in the attic
- Roughly where its temperature comes from
- It carries air below room temperature by design, and duct insulation only slows the surface toward the surrounding air.
- Below the 75°F dew point?
- Yes, whenever the system runs.
- What follows
- Building Science Corporation records condensation on ductwork and air handlers in vented attics as common in hot humid climates. This is usually the first thing to find and the cheapest thing to fix.
Air handler cabinet in the attic
- Roughly where its temperature comes from
- Same reason, with more surface area and more places for the insulation to be interrupted at a seam or a service panel.
- Below the 75°F dew point?
- Yes.
- What follows
- Water dripping off an air handler is frequently reported as a roof leak. It is not one, and no roofing work will change it.
Top face of the ceiling gypsum, under the insulation
- Roughly where its temperature comes from
- It sits close to the temperature of the air-conditioned room below it, at the thermostat setpoint.
- Below the 75°F dew point?
- Yes — 74°F is below 75°F, and the margin is one degree.
- What follows
- Any humid attic air that reaches it through a gap in the ceiling plane will wet it. It also shows why the thermostat setpoint is a moisture variable: the modelling reported alongside the DOE field study of this assembly showed high sensitivity to occupancy conditions.
A Class I vapor retarder on the room side of the ceiling
- Roughly where its temperature comes from
- Room temperature, and effectively closed to vapor at 0.1 perm or less.
- Below the 75°F dew point?
- Yes, and worse than yes.
- What follows
- It is a cold, impermeable face in the summer gradient. Vapor arriving from above stops there and cannot dry inward. BASC's warm-humid guidance is not to install one on the interior side of air-permeable insulation in the first place.
Interior gypsum behind a vinyl wallcovering
- Roughly where its temperature comes from
- Room temperature, with a low-permeance finish on the room side.
- Below the 75°F dew point?
- Yes.
- What follows
- Lstiburek's summary is blunt: vinyl wall coverings “are vapor barriers. Really. They are made out of vinyl,” with a Type II covering under 1 perm. It is a wall case rather than a roof case, and it is the clearest illustration on this page of a vapor retarder installed by accident.
Every surface temperature above is an assumption chosen to demonstrate the method. Substitute real numbers — measured with a surface thermometer and a hygrometer, or reported by whoever inspects the building — and the same three columns give you the answer for that building. The method transfers; the numbers do not.
What the audit just told you
Five of the seven surfaces in that table are not the roof at all. They are mechanical equipment, the ceiling plane, and layers somebody installed on the inside of the building — and two of the five, the ducts and the air handler, are cold surfaces sitting outside the thermal envelope in humid air, which is exactly why they collect water. That is the result this page exists to deliver. In a cooling-dominated climate the moisture problem that shows up on a roof deck usually originates with a cold surface that has nothing to do with roofing, and it is fixed by mechanical work, air sealing, and removing a layer somebody installed for the wrong climate.
It also tells you what a competent site visit looks like. Somebody with a surface thermometer, a hygrometer, and a moisture meter can fill in the third column for your building in an afternoon. Somebody who arrives with a product and no instruments is filling in a different form entirely.
What mold on the sheathing is telling you — and what it is notSection link
Growth on a roof deck records where a surface stayed wet. The pattern narrows the moisture source; it never proves one on its own. This table is a synthesis, not a source, and the third column is the discipline.
Everything below is read from photographs taken by a contractor or an energy auditor, not by the reader. Do not go up there. The pattern is visible in a photograph; the heat, the spores, and the ceiling joists are not.
| What the photographs show | What it narrows to | What it does not prove | What to do with it |
|---|---|---|---|
| A local dark patch directly above one bathroom or kitchen | An exhaust fan discharging into the attic, or a duct that has come apart. BASC's instruction is to vent fans outdoors, not into an attic, crawlspace, or space between floors. | That the fan is the only source. A house can have this and a general humidity problem at once. | Ask where every exhaust fan terminates, by fixture. It is cheap to check and cheap to fix, and it should be settled before anyone prices insulation. |
| Staining and rust that follow a duct run | Condensation on cold ductwork, which Building Science Corporation records as common in hot humid climates. | That the ducts are the whole load. Poor duct insulation, leaking connections, and a high attic dew point all contribute. | This is a mechanical scope item, not a roofing one. Ask what happens to the ductwork in any proposal, and whether it can be brought inside the envelope. |
| General darkening across the whole underside of the deck | The attic air itself is humid for long periods and the deck spends time below its dew point. | Which source is filling the attic — outdoor ventilation air, indoor air leaking up, or an unvented fan. All three look the same from below. | This is the case for a measurement rather than a guess. Attic temperature and humidity logged over a few weeks separates the possibilities. |
| A band of staining concentrated near the eaves | Humid outdoor ventilation air entering at the soffits and meeting cooler surfaces, the situation BASC describes as encouraging condensation in the attic. | That the venting is wrong. Vented attics work in humid climates when the ceiling is tight and nothing cold lives in the attic. | Raise the question of whether the assembly should be vented at all on this building, and read the ventilation guide before assuming more vents will help. |
| Growth that returns within a season of being cleaned | The moisture source was never fixed. EPA guidance is explicit that cleaning without fixing the water problem most likely brings the mold back. | That the remediation was done badly. It may have been done well and pointlessly. | This is the single most diagnostic observation on the page. Stop buying cleaning and start buying a moisture diagnosis. |
| Staining on stored contents but not on the deck | High attic humidity without the deck spending time below the dew point. Cardboard and fabric absorb and hold moisture that the wood does not. | That the assembly is safe. It means the deck has not been the cold surface, which can change with the seasons or with a mechanical change. | Treat it as an early warning. It is the cheapest moment to fix the humidity source that exists. |
| A defined path of staining that follows framing, after rain | Bulk water — a leak — rather than condensation. Water travels along framing before it appears. | That there is no condensation problem too. Both can run in the same attic. | This one is a roofing repair. Start with the leak page and a flashing inspection rather than with vapor strategy. |
Read this table one item at a time
A local dark patch directly above one bathroom or kitchen
- What it narrows to
- An exhaust fan discharging into the attic, or a duct that has come apart. BASC's instruction is to vent fans outdoors, not into an attic, crawlspace, or space between floors.
- What it does not prove
- That the fan is the only source. A house can have this and a general humidity problem at once.
- What to do with it
- Ask where every exhaust fan terminates, by fixture. It is cheap to check and cheap to fix, and it should be settled before anyone prices insulation.
Staining and rust that follow a duct run
- What it narrows to
- Condensation on cold ductwork, which Building Science Corporation records as common in hot humid climates.
- What it does not prove
- That the ducts are the whole load. Poor duct insulation, leaking connections, and a high attic dew point all contribute.
- What to do with it
- This is a mechanical scope item, not a roofing one. Ask what happens to the ductwork in any proposal, and whether it can be brought inside the envelope.
General darkening across the whole underside of the deck
- What it narrows to
- The attic air itself is humid for long periods and the deck spends time below its dew point.
- What it does not prove
- Which source is filling the attic — outdoor ventilation air, indoor air leaking up, or an unvented fan. All three look the same from below.
- What to do with it
- This is the case for a measurement rather than a guess. Attic temperature and humidity logged over a few weeks separates the possibilities.
A band of staining concentrated near the eaves
- What it narrows to
- Humid outdoor ventilation air entering at the soffits and meeting cooler surfaces, the situation BASC describes as encouraging condensation in the attic.
- What it does not prove
- That the venting is wrong. Vented attics work in humid climates when the ceiling is tight and nothing cold lives in the attic.
- What to do with it
- Raise the question of whether the assembly should be vented at all on this building, and read the ventilation guide before assuming more vents will help.
Growth that returns within a season of being cleaned
- What it narrows to
- The moisture source was never fixed. EPA guidance is explicit that cleaning without fixing the water problem most likely brings the mold back.
- What it does not prove
- That the remediation was done badly. It may have been done well and pointlessly.
- What to do with it
- This is the single most diagnostic observation on the page. Stop buying cleaning and start buying a moisture diagnosis.
Staining on stored contents but not on the deck
- What it narrows to
- High attic humidity without the deck spending time below the dew point. Cardboard and fabric absorb and hold moisture that the wood does not.
- What it does not prove
- That the assembly is safe. It means the deck has not been the cold surface, which can change with the seasons or with a mechanical change.
- What to do with it
- Treat it as an early warning. It is the cheapest moment to fix the humidity source that exists.
A defined path of staining that follows framing, after rain
- What it narrows to
- Bulk water — a leak — rather than condensation. Water travels along framing before it appears.
- What it does not prove
- That there is no condensation problem too. Both can run in the same attic.
- What to do with it
- This one is a roofing repair. Start with the leak page and a flashing inspection rather than with vapor strategy.
Cross-reference every row against the weather. Bulk-water intrusion correlates with rain and wind direction; condensation correlates with humidity and air-conditioner run time. If the moisture appears on hot, dry, sunny days, rain is not the cause.
What to change, in what orderSection link
The order matters more than any individual item, because several of these steps are hidden by the ones after them. Anything cheap that removes a moisture source comes before anything expensive that manages one.
Step one — stop adding water
Find every deliberate moisture source and remove it. Exhaust fans terminating in the attic, a soffit rather than the outdoors, or nowhere at all. Ducts with failed insulation or open connections. A condensate drain that has been backing up. An unsealed crawl space or an uncovered dirt floor beneath the house pushing vapor upward. BASC’s instruction on the first of those is direct: vent the fan outdoors, not into an attic, crawlspace, or space between floors, and insulate the duct to minimize condensation.
This is also the step where indoor humidity gets measured rather than assumed. The EPA target is below 60 percent relative humidity, ideally between 30 and 50. A hygrometer that costs less than a tank of fuel will tell you whether the building is meeting it.
Step two — stop moving air between the house and the attic
Air carries far more moisture than diffusion does, in either direction. In a vented attic with a reachable floor, sealing the air barrier at the ceiling plane — light fixtures, the attic hatch, plumbing stacks, duct chases, dropped soffits, the gap at the top plate — is the step that stops the exchange. It is the same work the ice dam page asks for in a cold climate, for the mirror-image reason, which is a useful reminder that the air barrier is climate-independent even when the vapor strategy is not. It is also work to buy rather than work to attempt: it happens in the attic, in the heat, over a ceiling that will not hold a foot, and it is done by an insulation or weatherization contractor.
Step three — decide which direction the assembly dries
Only now is it worth arguing about vapor retarders, ventilation, and unvented assemblies, because until the first two steps are done you are arguing about the smaller term. The decision has three honest outcomes, and which one applies is a building question:
- Keep it vented and let it dry outward and inward. Appropriate where the ceiling plane is tight, nothing cold lives in the attic, and there is no low-permeance layer on the room side. The cheapest correct answer when it is available.
- Remove the layer that is blocking inward drying. Where a Class I vapor retarder or a vinyl wallcovering was installed on the room side in a cooling-dominated climate, taking it out can be what lets the assembly recover.
- Go unvented and bring the equipment inside. The stronger answer where ductwork and the air handler are in the attic, and the one BASC points at for hot, humid locations. It is a design, not a product, and it costs you the leak visibility a vented attic gives you.
Step four — clean, last
Remediation is the final step, not the first, and its scope should say so. The EPA’s line is that the key to mold control is moisture control; cleaning ahead of the fix buys a delay rather than a result, and painting over it is worse than either — the EPA warns that paint applied over moldy surfaces is likely to peel. Past roughly ten square feet, that work belongs to somebody who does it for a living, and on a roof deck it is over your head the whole time.
What finishing looks like
A finished job has a moisture source named in writing, a cold surface either warmed or removed from the humid air, an assembly with a stated drying direction, a measured indoor humidity inside the EPA’s range, and a follow-up visit scheduled far enough out to cross a humid season. If any of those five is missing, the work is not finished; it is paused.
What changes this on a real buildingSection link
- Climate
“Hot-humid” has a technical definition, and it is worth knowing because it is not the same as “the South.” The Department of Energy’s Building America program defines a hot-humid climate as a region receiving “more than 20 in. (50 cm) of annual precipitation” where one or both of the following occur: a wet-bulb temperature of 67 °F or above for 3,000 or more hours during the warmest six consecutive months, or 73 °F or above for 1,500 or more hours over the same period.
Mixed-humid is defined separately — more than 20 in of precipitation, roughly 5,400 heating degree days (65 °F basis) or fewer, and average monthly outdoor temperatures dropping below 45 °F in winter — and it is the hardest case on this page, because the vapor drive genuinely reverses. An assembly there has to be able to dry in both directions, which usually means avoiding a low-permeance layer on either face rather than choosing which face to put one on.
A climate zone is a design input, not a diagnosis. Two houses in the same zone with different mechanical layouts, different interior humidity, and different thermostat habits will not behave the same way.- Moisture and ventilation
Before any of the roof questions, count the moisture sources inside the building. A large one is often installed deliberately: BASC’s instruction for exhaust fans is to “Install the fan to vent outdoors, not into an attic, crawlspace, or space between floors,” and adds that “To minimize condensation, insulate the duct.” A bath fan discharging into an attic is a humidifier aimed at the underside of the deck, twice a day, every day.
The EPA’s target for indoor air is to “Keep indoor humidity below 60 percent (ideally between 30 and 50 percent) relative humidity.” In a hot-humid climate that is a mechanical question — an air conditioner sized so large that it satisfies the thermostat before it dehumidifies will leave the house at a high indoor humidity while feeling cold. A roofing contractor cannot see that from the roof.
- Code and jurisdiction
There is no nationwide building code for site-built houses in the United States. Every provision quoted on this page is model text as summarized in a U.S. Department of Energy code-compliance brief — language a jurisdiction may adopt, amend, delay, or decline. This page does not tell you what applies at your address, and it does not pin the provisions to an edition, because DOE’s brief is written against the 2015 IRC while attributing the diffusion-port language to 2018 ICC proposals.
Separately: an unvented conversion normally touches the energy code, the fire code, and the mechanical code at once, and whether a re-roof or an insulation retrofit triggers compliance is itself a local question with a local answer.
Record the jurisdiction, the adopted edition, the amendments, the effective date, and the official URL for any code claim, and confirm with the authority having jurisdiction. Nothing on this page is a code determination for your building.- Ventilation — direction-dependent, not universally good
In a heating climate ventilation air is a drying agent. In a cooling-dominated one it is the moisture supply, which is why BASC describes humid ventilation air meeting cooler inside surfaces as “encouraging condensation in the attic.” That does not mean vented attics are wrong in the South; a vented attic with a tight ceiling, no ducts above it, and no interior vapor barrier works in a great many humid-climate houses. It means the vent area is not the variable that decides it.
There is no universal ventilation ratio, and “more ventilation is always better” is false. Required net free area, the conditions attached to it, and whether it applies to your assembly at all are set by the code edition your jurisdiction adopted, the climate zone, and the assembly type. Confirm with the authority having jurisdiction before designing to any number, including one from this site.- Unvented assemblies — legitimate, and easy to do badly
Vented and correctly designed unvented assemblies are both legitimate, and in a hot-humid climate with mechanical equipment in the attic the unvented version is often the better engineering answer. The failure mode is not the concept. It is the improvised version: a few inches of foam sprayed under an old deck, the soffit vents left open or blocked at random, no decision made about which way the assembly is expected to dry.
The most recent DOE Building America field work on the hot-humid version of this assembly — five occupied new homes in DeBary, Florida, IECC climate zone 2A, monitored over nine months — found “no signs of sustained mold, rot, or corrosion risk” and “no evidence of condensation dripping from ductwork.” It also found “high sensitivity to occupancy conditions” in modelling, and its authors concluded that diffusion ports “should not be widely recommended as the sole method to mitigate attic moisture issues” without further investigation. Five houses over nine months is real evidence and it is a small sample; both of those things are true at once.
Converting a vented attic to an unvented one is a design decision with vapor-control, code, fire, mechanical, and warranty consequences. It is answered for a specific building by a qualified designer, not by a table or a spray-foam quote.- Fire
Two fire points attach to this work. First, a roof’s Class A, B, or C classification is a property of a tested assembly — deck, underlayment, and covering together — not of the visible covering alone, so changing the underlayment or adding insulation above the deck can change what assembly you actually have. Second, foam plastic insulation sprayed under a deck is regulated for ignition and thermal barriers, and whether a particular product may be left exposed in a particular attic is set by its listing and by the adopted code, not by the installer.
Fire classification applies to a tested assembly. Ask which listed assembly the finished roof matches, and ask for the ignition-barrier or thermal-barrier basis for any exposed foam in writing.- Structural weight
Persistent wetting is a structural question eventually, not just a cosmetic one. Decay fungi attack the wood itself rather than growing on the surface of it, and sheathing or framing that has been repeatedly wet can lose capacity while still looking acceptable in a photograph. This page cannot tell you whether a given deck is sound; that is a determination made on the building, by someone qualified, with a probe and a moisture meter, and if anything is deflecting or sagging it is an evacuate-and-call situation rather than a maintenance one.
- Maintenance
The maintenance obligation in a humid climate is mostly mechanical, not roofing. Filters, condensate drains, duct connections, and dehumidifier operation all decide whether the roof assembly sees a moisture load. A condensate pan overflowing into an attic is the same water stain on a ceiling as a roof leak and it is not a roof leak, which is one of the more expensive misdiagnoses in this category.
- Access and site conditions
Everything on this page can be established without a reader entering an attic: photographs from a contractor or energy auditor, moisture-meter readings taken by someone with the right equipment, a blower-door and infrared report, and the ceiling side of the assembly. In a cooling climate an attic is also a heat hazard in its own right, which is the subject of the callout at the top of this page.
Do not enter an attic to confirm anything here. Heat, disturbed insulation, airborne spores, and a foot through a ceiling are all more likely than a useful finding.- Health, insurance, and disclosure
This page is not medical advice and does not evaluate health effects. The EPA’s own instruction is to “consult a health professional before starting cleanup” if you have health concerns, and that is the right route for anyone in the household with asthma, a lung condition, or a suppressed immune system. On money: whether a policy responds to mold or to condensation damage, what it treats as the covered cause, how sub-limits and exclusions apply, and what a seller must disclose are governed by the specific policy, the facts, and the law of the state involved. Nothing here predicts an outcome, and any page that tells you that you are covered for mold is guessing.
Coverage, deductibles, sub-limits, claim outcomes, and disclosure duties vary by carrier, state, and facts. Read the policy and, where money or a sale is at stake, take advice from someone licensed in your state.
Which documents this work puts in playSection link
Changing how a roof assembly manages moisture touches more paperwork than most people expect, and the documents cover different things. None of them is a substitute for reading it.
- The covering manufacturer's limited warranty
Some steep-slope covering warranties attach conditions to attic ventilation, and some address unvented assemblies specifically. This page cannot tell you which, because that is set by the individual document for the individual product. The action item is concrete: before anyone sprays foam against the deck or closes a soffit vent, get the manufacturer’s current published position on unvented assemblies for that exact product, in writing, and read what it says about ventilation as a condition of coverage.
- The insulation product's own instructions
Spray foam and rigid board are separate products with separate instructions covering substrate condition, thickness per pass, cure, ignition and thermal barriers, and permissible exposure. Those instructions are product-specific, and they are the document a code official will ask for.
- The installer's workmanship warranty
Read it for length, for what triggers a callback, for whether the diagnostic visit is chargeable, for whether it survives a sale, and for what happens if the company stops trading. On moisture work specifically, ask what the warranty says if the mold returns — because if the moisture source was never identified, it will.
- The remediation contractor's scope
A remediation scope that ends at cleaning is a scope that ends before the problem does. The EPA’s framing is that “the key to mold control is moisture control,” so the document should name the moisture source, name who is fixing it, and say what happens if cleaning happens and the fix does not.
- Insurance is a separate document, not a warranty
Mold and condensation are among the most heavily conditioned areas of residential property coverage, and they vary by carrier, by state, and by the facts of the loss. This page does not tell you what your policy does. Read the policy, and where the amounts are significant, take advice from someone licensed where you live.
Repairability
What this costs to put right depends almost entirely on how far the water got and on whether the source is fixable without opening the assembly. The order is fixed: nobody should be cleaning, coating, or refinishing anything before the moisture source is identified and stopped.
- Mechanical fixes: re-ducting a bath fan to the outdoors, insulating a duct, sealing duct connections, correcting a condensate drain. Usually the cheapest work on the list and frequently the whole answer.
- Ceiling-plane air sealing: reachable in a vented attic with an accessible floor, and the step that stops humid air being pumped between the house and the attic in either direction.
- Removing an interior vapor retarder: genuinely disruptive, because it is usually behind a finished ceiling or a wallcovering. In a cooling-dominated climate it can still be the thing that lets the assembly dry.
- Surface mold on sheathing: a cleaning question at small scale and a professional remediation question past the EPA’s roughly 10-square-foot line. Repeat growth after cleaning is diagnostic information, not bad luck.
- Deck replacement: the point at which this stops being a moisture project and becomes a roofing project, and it is normally done at re-roof because the covering has to come off anyway.
- Converting to an unvented assembly: effectively unrepairable in stages. It is a design decision made once, and the sensible moment for it is when the roof is being replaced.
A warranty is a contract between a reader and whoever wrote it. What it covers, what voids it, whether it transfers, and how it is enforced are set by that document and by the law where the reader lives. Read the actual warranty for the product and the installer in front of you — not a summary of one, including this one.
Questions to ask an installerSection link
Moisture in a humid climate draws roofers, insulation contractors, HVAC contractors, and remediation firms, and each of them will tend to find the problem their trade solves. These questions are designed to reveal which problem the person in front of you thinks they are solving.
Which way do you think vapor is driving through this assembly, and in which season?
It is the first question and it is diagnostic on its own. Anyone who answers with a direction and a season is thinking about the building. Anyone who answers with a product is selling one.
Which surface in this assembly do you believe is below the dew point, and what is that surface at?
Condensation happens on a specific surface, and it can be measured. A good answer names one — the ducts, the ceiling, the deck at night — and says how it was established.
Is there a vapor retarder anywhere in this assembly now? What class, and on which face?
In a warm-humid climate the answer can be the whole problem. BASC is explicit that Class I vapor retarders should not be installed on the interior side of air-permeable insulation there, and existing ones are often invisible until someone opens the ceiling.
Where do the bathroom and kitchen exhaust fans actually terminate?
BASC’s instruction is to vent them outdoors and not into an attic, crawlspace, or space between floors. If they terminate in the attic, that is a fixable, deliberate moisture source and it should be near the top of any scope.
Is there ductwork or an air handler in the attic, and what is the plan for it?
Cold equipment in humid air is the most common condensing surface in a hot-humid attic. If the answer is that it stays exactly as it is, the proposal is treating the symptom.
If you are proposing an unvented attic, in which direction is the assembly designed to dry?
Every assembly needs a drying direction. If the answer is “it will not get wet,” the answer is that there is no plan for the day it does.
If you are proposing a vapor diffusion port, what is its area, its permeance, and where does it sit relative to the ridge?
The model provisions are specific: not more than 12 inches from the highest point of the roof, area at least 1:600 of the ceiling area, and a membrane of at least 20 perms tested to ASTM E96 Procedure A. Someone proposing the detail should know its three numbers without looking.
What happens to the roof covering warranty and the workmanship warranty if the attic is sealed?
Ask for the manufacturer’s current published position on unvented assemblies for the specific product on the roof, in writing, before the work starts rather than after.
Before cleaning any mold, what is the moisture source, and who is fixing it?
EPA guidance is that the key to mold control is moisture control, and that cleaning without fixing the water problem most likely brings the mold back. A remediation quote with no moisture-source sentence is a quote for doing this twice.
Was this house built before 1990, and how are you handling old felts, mastics, and duct or pipe insulation if you find them?
Older roofing and insulation materials may contain asbestos. The EPA recommends testing suspect materials where a planned renovation would disturb them, says samples “should be taken by a properly trained and accredited asbestos professional (inspector),” and states that laboratory testing is the only way to be sure. The correct answer here is that protocol, not reassurance.
Require these in writing
- The diagnosis in one sentence, naming the moisture source and the condensing surface the work is intended to remove.
- The climate zone used, and which vapor strategy the design is built around.
- Every existing vapor retarder found, with its class and its face, and what happens to it.
- Exhaust fan terminations, before and after, by fixture.
- The location of ductwork and the air handler after the work, and whether they end up inside the thermal envelope.
- For an unvented proposal: insulation type, thickness, placement relative to the deck, and the stated drying direction.
- For a diffusion port: area as a ratio of ceiling area, membrane permeance and test method, and distance from the highest point of the roof.
- Ignition-barrier or thermal-barrier basis for any exposed foam, cited to the product listing.
- Whether the work triggers a permit in your jurisdiction, and who pulls it.
- Effect on the covering warranty and the workmanship warranty, in writing, from the manufacturer rather than from the salesperson.
- A pre-1990 asbestos protocol if the building qualifies.
Misconceptions and failure modesSection link
Common misconceptions
Common belief
Vapor barriers keep moisture out, so more of them is safer.
What is actually true
A low-permeance layer stops vapor in one direction and stops drying in the other. Whether that is protection or a trap depends entirely on which face it is on relative to the season’s vapor drive. BASC’s warm-humid guidance is a prohibition: do not install Class I vapor retarders on the interior side of air-permeable insulation in above-grade walls, except at shower and tub walls. Two barriers, one on each face, is the worst case of all — an assembly that can never dry.
Common belief
Condensation in a roof is a cold-weather problem.
What is actually true
It is a cold-surface problem, and in an air-conditioned building in a humid climate the cold surfaces are indoors in August. Lstiburek records condensation on ductwork and air handlers in vented attics as common in hot humid climates. The calendar is irrelevant; the temperature of the surface relative to the dew point of the air touching it is not.
Common belief
More attic ventilation will dry it out.
What is actually true
In a cold climate, yes, because the ventilation air is dry. In a hot-humid climate the ventilation air is the moisture supply — BASC describes humid outdoor air introduced into a vented attic meeting cooler inside surfaces as encouraging condensation there. There is no universal ventilation ratio, and both vented and correctly designed unvented assemblies are legitimate.
Common belief
Mold on the roof sheathing means the roof has failed.
What is actually true
Usually it means air and vapor reached a cold surface and stayed there, which is a building-science outcome rather than a roofing defect. A brand-new covering over the same assembly will produce the same staining. What a re-roof genuinely buys is the chance to correct the assembly while the covering is off — a different and better argument than “the roof failed.”
Common belief
Kill the mold and the problem is solved.
What is actually true
EPA guidance is that the key to mold control is moisture control, and that if you clean up the mold but do not fix the water problem the mold problem will most likely come back. The EPA also says plainly not to paint or caulk over moldy surfaces, because paint applied over them is likely to peel. Cleaning is the second step.
Common belief
Spray foam under the deck fixes humidity problems.
What is actually true
A correctly designed unvented assembly can be exactly the right answer in a hot-humid climate, particularly where the ductwork and air handler are in the attic. But foam is a component, not a design. The questions that decide it — which direction the assembly dries, whether the insulation is air-impermeable, whether the space is inside the thermal envelope, what the local code requires, and what it does to the covering warranty — are not answered by the presence of foam.
Common belief
A vapor diffusion port is a proven fix for humid attics.
What is actually true
It is a real detail with model-code recognition in the hot climate zones and it is genuinely promising, and the DOE Building America field study of it in Florida found no signs of sustained mold, rot, or corrosion risk across the homes monitored. The same study concluded that diffusion ports “should not be widely recommended as the sole method to mitigate attic moisture issues” pending further investigation. Both halves belong in the sentence.
Common belief
A wet ceiling stain in summer means there is a roof leak.
What is actually true
In a humid climate it might equally be a blocked condensate drain, a sweating duct, or condensation on a chilled surface. The distinguishing evidence is the weather correlation: bulk water intrusion tracks rain, and condensation tracks humidity and air-conditioner run time. If the stain appears on hot, dry, sunny days, rain is not the cause.
How it actually fails
- Cold-climate detailing built in a cooling climate
- A polyethylene sheet or foil-faced product installed on the room side of the ceiling, on the pattern of guidance written for a heating climate. In summer that layer is on the cold face, so inward-driven vapor stops at it and the assembly cannot dry inward. BASC’s warm-humid guidance is not to install Class I vapor retarders there in the first place.What you can see: Staining or damp on the attic side of the ceiling rather than on the deck. Musty smell that peaks in the humid season. Wet insulation lying directly on the ceiling.
- Sweating ducts in a vented attic
- Supply ducts and the air handler run below room temperature, and the vented attic delivers outdoor air that is well above their dew point. Lstiburek records this as common in hot humid climates. Insufficient duct insulation and unsealed connections both make it worse.What you can see: Water dripping from ducts or the air handler when it has not rained. Rust at duct straps and hangers. Ceiling stains that follow duct runs rather than roof geometry.
- Bath or kitchen fan discharging into the attic
- The duct terminates in the attic, into a soffit, or nowhere at all. BASC’s instruction is to install the fan to vent outdoors, not into an attic, crawlspace, or space between floors, and to insulate the duct to minimize condensation.What you can see: A local patch of dark staining on the deck directly above a bathroom or kitchen. A visible flexible duct in the attic that ends in mid-air or disappears into insulation.
- Improvised unvented attic
- Foam sprayed under an old deck without a decision about drying direction, with soffit vents partly open, partly blocked, and no answer to what happens when the deck gets wet. The concept is legitimate; this execution is a sealed assembly with no exit.What you can see: Foam applied only between rafters with gaps at the eaves. Soffit vents still visible from the ground under a supposedly sealed attic. No documentation of insulation type, thickness, or code pathway.
- Remediation without a moisture fix
- The visible growth is cleaned or encapsulated and the source is never named. EPA guidance is that cleaning without fixing the water problem most likely brings the mold back, and that painting over moldy surfaces is likely to peel.What you can see: Growth returning in the same outline within a season or two. Peeling coating on treated sheathing. A remediation invoice with no moisture-source line on it.
- Oversized air conditioning leaving the house humid
- A system large enough to satisfy the thermostat quickly runs in short cycles and removes less moisture, so the house is cold and still humid. That raises the indoor dew point above the EPA’s recommended range and loads every cool surface in the building.What you can see: A house that feels clammy at the set temperature. Condensation on supply registers and on window glass in summer. Indoor relative humidity readings persistently above 60 percent.
- Roof leak misread as condensation, or the reverse
- Both produce wet sheathing and a ceiling stain, and the trades involved have opposite instincts. A leak is bulk water and tracks rain; condensation tracks humidity and equipment run time.What you can see: For a leak: a defined path, staining that follows framing, timing that matches wind-driven rain. For condensation: general staining across bays, no rain correlation, and moisture that is worst in the humid season.
- Assembly with no drying direction at all
- A low-permeance layer on the interior and another above the deck — for example an interior vapor barrier plus an impermeable underlayment or coating — so any moisture that gets in by air leakage or a small leak cannot leave in either direction.What you can see: Sheathing that stays wet long after the weather event that wet it. Repeat staining with no identifiable current leak. Moisture-meter readings that do not fall between visits.
Sources and further readingSection link
Understanding Roofing / Published
Scope and limitations
- It cannot tell you which surface in your building is below the dew point.
- That is a measurement on that building, on a day, with an instrument — not something any page can infer from a climate zone.
- It cannot tell you whether a stain is a leak or condensation.
- The two look identical from below and are distinguished by weather correlation, timing, and evidence gathered on site.
- It does not pin the model code provisions to an edition.
- The Department of Energy brief this page relies on is written against the 2015 IRC while attributing the vapor-diffusion-port language to 2018 ICC proposals, and this site did not read the ICC's own published text.
- Only the edition your jurisdiction adopted, with its amendments, governs anything.
- It is not a code determination, and it is not a design.
- Converting a vented attic to an unvented one, or specifying a vapor retarder, is engineering for a specific building and belongs to a qualified professional.
- It says nothing about the health effects of mold exposure, and it is not medical advice.
- Health questions go to a health professional.
- It cannot tell you whether an insurance policy responds to mold or condensation damage, or what a seller must disclose.
- Those are set by the policy, the facts, and state law.
- It contains no cost figures, because this site has no sourced cost data for moisture retrofits, unvented conversions, or remediation that would carry a scope, geography, date, and confidence statement.
- A range invented to fill the gap would be worse than the gap.
- The field evidence for the hot-humid unvented assembly it describes most favourably is five monitored houses in one Florida town over nine months.
- That is real and it is small.
Vented versus Unvented Attic
U.S. Department of Energy, Building America Solution Center (PNNL)
That in hot, humid locations the outside air is likely to be more humid than the inside air, that inside surfaces may be cooler and may provide condensing surfaces for humid air introduced into a vented attic, thus encouraging condensation in the attic; that in hot, humid locations it may be advisable to construct unvented rather than vented attics to help reduce the entry of moisture-laden air; that where HVAC equipment is in the attic an unvented assembly will be much more efficient because the ductwork and air handler will be inside the thermal envelope; that a vented attic insulated at the attic floor makes it easier to detect and solve roof leaks or other moisture-related issues and that exposed, ventilated rafters and decking can dry out more readily after a leak; and that the choice turns on climate, desire for additional living and storage space, building design and configuration, and location of HVAC.
Best-practice guidance for builders and designers, not adopted law. It does not state a ventilation ratio and it is not a code determination anywhere. It does not evaluate any specific product.
Class I Vapor Retarders Not Installed in Above-Grade Walls in Warm-Humid Climate
U.S. Department of Energy, Building America Solution Center (PNNL)
That in warm-humid climates one should not install Class 1 vapor retarders on the interior side of air-permeable insulation in above-grade walls, except at shower and tub walls; and that during summer months, when air conditioning is prevalent, condensation can also occur inside the home in localized areas where warm air comes in contact with cool surfaces.
Written for above-grade walls, not for roof assemblies. This page uses it for the direction-of-drive principle and for the interior-vapor-retarder prohibition, and does not extend its specific wall requirement to roofs. It is guidance, not adopted code.
Unvented Attic Insulation
U.S. Department of Energy, Building America Solution Center (PNNL)
That unvented roof assemblies are created by eliminating ventilation openings and moving the thermal, moisture, and air control boundaries to the plane of the roof deck; and that the key to creating an unvented roof assembly is to keep the roof deck — the principle condensing surface in roof assemblies — sufficiently warm throughout the year such that condensation will not occur.
Framed primarily around heating-climate condensation control. Its code references are to specific IRC and IECC editions and are model provisions, not the law in any jurisdiction.
Controlling Moisture in Unvented Attics — Code Compliance Brief
U.S. Department of Energy, Building America Solution Center (PNNL) / Written against the 2015 IRC and IECC, with vapor-diffusion-port language attributed to 2018 ICC code proposals
The MODEL unvented-attic provisions quoted on this page: that in climate zones 5, 6, 7 and 8 any air-impermeable insulation shall be a Class II vapor retarder or shall have a Class II vapor retarder coating or covering in direct contact with its underside; that where air-permeable insulation is installed below the deck, rigid board or sheet insulation above the structural roof sheathing may alternatively be sized to maintain the monthly average temperature of the underside of that sheathing above 45°F; that the unvented space must be completely within the building thermal envelope and that no interior Class I vapor retarder is installed on the ceiling side of the assembly; and, for air-permeable insulation in climate zones 1, 2 and 3, that an approved vapor diffusion port shall be installed not more than 12 inches (305 mm) from the highest point of the roof, with a port area ≥1:600 of the ceiling area and a vapor permeable membrane rated ≥20 perms when tested in accordance with Procedure A of ASTM E96.
This is DOE's summary of MODEL code text, not adopted law, and not the International Code Council's own publication — this site could not access the ICC text, which returned HTTP 403 during drafting, and therefore does not pin these provisions to a published edition or section number. Model text has no jurisdiction and no effective date of its own; it becomes law only where a government adopts it, on that government's effective date, and adoptions routinely amend it. Confirm the adopted edition, its amendments, and its effective date with your authority having jurisdiction.
Bathroom Exhaust Fans
U.S. Department of Energy, Building America Solution Center (PNNL)
That the fan should be installed to vent outdoors, not into an attic, crawlspace, or space between floors; and that to minimize condensation, the duct should be insulated.
Guidance for bathroom exhaust fans specifically. It does not quantify the moisture load an attic-terminated fan adds, and it is not adopted code anywhere.
Climate Zones
U.S. Department of Energy, Building America program
The definition of a hot-humid climate — a region receiving more than 20 in. (50 cm) of annual precipitation where a wet-bulb temperature of 67°F or higher occurs for 3,000 or more hours, or 73°F or higher for 1,500 or more hours, during the warmest six consecutive months — and the mixed-humid definition of more than 20 in. of annual precipitation, approximately 5,400 heating degree days (65°F basis) or fewer, and average monthly outdoor temperatures dropping below 45°F during winter.
Defines Building America climate regions. The page as read does not map those regions onto specific IECC zone numbers, so this page does not claim a zone-number equivalence beyond the zones the DOE code brief itself names.
A Brief Guide to Mold, Moisture and Your Home
U.S. Environmental Protection Agency
That the key to mold control is moisture control; that cleaning up mold without fixing the water problem most likely means the mold problem will come back; and that indoor humidity should be kept below 60 percent, ideally between 30 and 50 percent, relative humidity.
General homeowner guidance about indoor mold. It is not roofing guidance, it does not address roof assemblies, and it does not evaluate health effects for any individual.
Mold Cleanup in Your Home
U.S. Environmental Protection Agency
That if the moldy area is less than about 10 square feet — less than roughly a 3 ft by 3 ft patch — in most cases a homeowner can handle the job themselves; that one should not paint or caulk moldy surfaces because paint applied over them is likely to peel; that plumbing leaks and other water problems should be fixed as soon as possible and all items dried completely; and that anyone with health concerns should consult a health professional before starting cleanup.
Homeowner cleanup guidance. The 10-square-foot line is a rule of thumb in an EPA consumer document, not a regulatory threshold, and it says nothing about work over your head on a roof deck.
Why Dew Point Is More Important Than Relative Humidity
National Weather Service, La Crosse, Wisconsin
The definition of dew point as the temperature the air needs to be cooled to, at constant pressure, in order to achieve a relative humidity of 100%; and the summer comfort bands of 55°F or below as dry and comfortable, 55–65°F as becoming sticky, and 65°F and above as lots of moisture in the air, becoming oppressive.
A public-education page about outdoor comfort. Its comfort bands describe how air feels to people, not what an assembly will do. The 75°F dew point used in this page's worked example is an assumption for arithmetic, not a value taken from this source.
Heat — Health Effects
U.S. Occupational Safety and Health Administration
That hazardous heat exposure can occur indoors or outdoors and during any season if the conditions are right, not only during heat waves; and that heat stroke — signalled by mental dysfunction such as unconsciousness, confusion, disorientation, or slurred speech — requires immediate medical attention, with the instruction to cool the person immediately and call 911.
Occupational-safety guidance for employers and workers. It is not homeowner guidance and it does not address attics specifically; it is cited here for the nature and urgency of the hazard, not as a procedure to copy.
BA-2401: Moisture Performance of Unvented Attics with Vapor Diffusion Ports and Buried Ducts in Hot, Humid Climates
Kohta Ueno, Building Science Corporation, for the U.S. Department of Energy Building America program / 21 September 2024
That five occupied new-construction homes in DeBary, Florida (IECC climate zone 2A) were monitored over nine months at the roof deck and attic floor; that field observations showed no signs of sustained mold, rot, or corrosion risk and no evidence of condensation dripping from ductwork; that modelling under stressful conditions showed high sensitivity to occupancy conditions; and that the authors concluded vapor diffusion ports should not be widely recommended as the sole method to mitigate attic moisture issues without further investigation.
A research report from a DOE program, published by the research contractor rather than by the department. Five houses in one Florida town over nine months is a small sample and a short window, and the report's own conclusion is explicitly provisional. It is not a code pathway and not a recommendation for any specific building.
BSD-102: Understanding Attic Ventilation
Joseph Lstiburek, Building Science Corporation / 25 October 2006
That in hot humid climates condensation on ductwork and air handlers located in vented attics is common; and that unvented attic and roof designs have the advantage of providing conditioned spaces for ductwork and air handlers.
A building-science digest from 2006, written for practitioners. It is not adopted code, its code references predate current editions, and its conclusions are the author's professional position rather than a research finding.
BSI-130: Hotel Mold
Joseph Lstiburek, Building Science Corporation / 14 March 2022
That vinyl wall coverings are vapor barriers, with a Type II (20 oz) covering measuring less than 1 perm; and the general position that if you air-condition and you have humid summers there should be no interior vapor barrier.
A building-science opinion article about commercial hotel walls, not about residential roof assemblies. This page uses it for the interior-vapor-barrier principle and for the permeance of vinyl wallcoverings, not for any roof-specific requirement.
How do I know if I have asbestos in my home (floor tile, ceiling tile, shingles, siding, etc.)?
U.S. Environmental Protection Agency
That EPA recommends testing suspect materials if they are damaged or if a planned renovation would disturb them; that samples should be taken by a properly trained and accredited asbestos professional (inspector); and that the only way to be sure whether a material contains asbestos is to have it tested by a qualified laboratory.
General homeowner guidance. It does not identify which specific roofing or insulation products contain asbestos, and state and local rules on testing, notification, and disposal vary.
Code Update on Vapor Diffusion Ports: What does it mean for Georgians?
Southface Institute / 29 July 2025
The single narrow claim that a jurisdiction amended this exact provision: that vapor diffusion ports have been in the IRC since 2018 at Section R806.5 Item 5.2, that they are allowed in climate zones 1, 2 and 3, and that Georgia deleted the language when it adopted the 2018 IRC in 2020 and reinstated it at a later code review.
A secondary source — a nonprofit institute's blog post. This site attempted to confirm the amendment history against the Georgia Department of Community Affairs' own published IRC amendment document and could not extract the relevant text, so this account is not treated as a jurisdiction determination and is used only as an illustration that adopted text and model text differ. It is not sole support for any technical or safety claim on this page.