No. Neither open-cell nor closed-cell spray foam rots a roof deck by itself, because foam is not water and wood does not rot unless it stays wet. What spray foam does change is how the deck dries and, more importantly, how you find out that water got in at all.
That distinction matters, and most of the internet skips it. The scary stories you have read are real, but when you read the details, the cause is almost always rainwater getting past the roof covering or indoor moisture condensing on cold sheathing. The foam did not eat the wood. It made the problem quiet.
The short version
Foam sprayed to the underside of your sheathing hides a leak that used to show up as a ceiling stain. That is the real risk, and it is a risk about inspectability, not chemistry. Fix the roof first, know how many years your roof covering has left, and in our marine climate use closed-cell for this assembly rather than open-cell.
Where the fear actually comes from
It comes from real houses. Homeowners have opened up a foamed roof and found black, soft sheathing. Building science writers have argued about it in public for years, which is why searching this question drops you into a fight between people who all know what they are talking about.
Here is the honest summary of that fight. Reported roof deck failures under spray foam have been associated with two things: bulk water leaking in from outside, and moisture from inside the house condensing on cold sheathing. Nobody credible has shown that cured foam spontaneously destroys the wood it is bonded to. So the useful question is not “does foam rot wood.” The useful question is “what does foam change about how water gets to this deck, sits on it, and gets noticed.”
Three answers to that, in order of how much they should worry you.
1. Foam hides a leak. This is the big one.
In a normal vented attic, a leak announces itself. Water runs down the rafter, drips onto the attic floor, and eventually shows up as a brown ring on somebody’s bedroom ceiling. You get annoyed, you call a roofer, and the deck dries out.
Bond foam to the underside of that sheathing and the announcement stops. Water that gets past the shingles now sits in the gap between the roofing and the wood, or tracks sideways along the deck, with insulation pressed against the bottom of it. There is no drip, no stain, and no smell until it is bad. A slow leak that would have cost you a few hundred dollars in a vented attic can quietly work on the deck for a couple of seasons before anyone knows.
This is not a chemistry problem and it is not solved by picking a different foam. It is a tradeoff you accept knowingly: you are buying real performance and giving up easy inspection. That trade is fine on a sound roof with life left in it. It is a bad trade on a roof that is nearly done.
2. Open-cell lets indoor moisture reach the cold sheathing
Open-cell foam is vapor open. It air seals, but water vapor can diffuse through it. Under a roof deck that means the humidity your household generates, from showers, cooking, laundry, and people breathing, can move through the insulation and reach the back of sheathing that is cold because it is one layer away from a wet Seattle night.
When a surface is below the dew point of the air touching it, moisture condenses on it. Do that often enough and the sheathing stays damp. There are documented cases of damp or wet OSB under open-cell roof applications, and they are not urban legend.
Closed-cell behaves differently in this assembly. At sufficient thickness it is its own vapor retarder, so indoor moisture does not diffuse through to the sheathing, and it does not absorb and hold water the way open-cell does. It also keeps the underside of the deck warmer, which raises the surface temperature above the dew point instead of leaving it cold. That is the main reason we prefer closed-cell for roof decks in the Puget Sound.
3. Drying, and why it cuts both ways
Once foam is on, that deck mostly dries in one direction: outward, through the roof covering, which is slow. Under most roofing that is fine, because the amount of water involved is small. It stops being fine when the source is continuous, like an actual leak. The foam did not cause the water. It just removed the fast drying path and the early warning.
| At the roof deck | Open-cell | Closed-cell |
|---|---|---|
| Lets indoor vapor through | Yes, it stays vapor open | No, at sufficient thickness |
| Absorbs and holds water | Yes | Resists it |
| Keeps sheathing warmer | Less, per inch | More, per inch |
| Condensation risk on sheathing | Higher, this is the documented concern | Lower |
| Hides a roof leak | Yes | Yes, equally |
| Our default here | Rarely on a roof deck | Usually |
Notice the last two rows. Closed-cell solves the condensation half of this and does nothing at all about the hidden leak half. Anybody who tells you closed-cell makes the roof risk go away is selling. For the full comparison, read open-cell vs closed-cell spray foam.
What the code says, and what that means in Seattle
You will see this come up in every online argument, usually stated with more confidence than it deserves, so here is the careful version.
In colder climate zones, generally IECC zone 5 and above, codes typically require a vapor retarder when open-cell foam is applied to the underside of a roof deck in an unvented assembly. That requirement exists precisely because of the condensation issue described above, and in a cold zone it is a sensible rule.
Seattle is not in that group. We are climate zone 4C, marine, which is a milder and wetter animal than zone 5 and the vapor retarder requirement does not apply here the same way. That is context, not a permission slip. Code adoption, amendments, and how a given assembly gets reviewed vary, and the requirement for your specific job is something we confirm with your local building department or inspector before we spray, not something we or anyone else should assert to you off a web page. We are not going to state a specific Washington requirement here as fact. We will get you the actual answer for your address.
Our practical position is simpler than the code question anyway. In a 4C marine climate we prefer closed-cell on roof decks, at proper thickness, on a roof that is watertight. That choice sidesteps most of the vapor argument on its own.
How to do this without creating a problem
Fix the roof first
Any active leak, any flashing you are unsure about, any suspect valley. Deal with it before insulation goes anywhere near the underside of that deck.
Know the roof’s remaining life
If the covering has five years left, insulating under it now means paying twice, and living with a hidden deck in the meantime. Reroof first, then foam.
Inspect the deck while you can see it
We look at the sheathing before we spray. Existing staining, soft spots, delaminated OSB, or old repair patches change the plan.
Spray it correctly
Right product for the assembly, right thickness, correct substrate temperature, controlled passes, full coverage with no gaps at rafter edges or the ridge.
Detailing is where this is won or lost. Gaps, thin spots, and missed edges create cold surfaces exactly where moist air can reach them, which is the condensation problem reintroduced by sloppiness. That is the same argument we make everywhere on this site: what a quality spray foam install looks like is not a marketing page, it is the actual risk control.
What to watch for afterwards
You have traded easy inspection for performance, so check on it deliberately once or twice a year. It takes ten minutes.
- Staining at the eaves and rakes. The edges are usually the first place water finds a way around the foam and becomes visible.
- A musty or earthy smell in the attic. Your nose finds wet wood before your eyes do. If the space smells different than it did last year, call someone.
- Sagging or waviness in the roof plane. Look up at the roof from the street in raking light. A dip between rafters that was not there before is worth investigating.
- Soft spots underfoot for your roofer. Mention the foam to anyone who goes up there, and ask them to tell you how the deck felt.
- Foam that pulled away from a rafter. A visible gap at the edge of a bay can mean adhesion trouble or movement, and it is a cold surface where there should not be one.
Wet sheathing does not go from fine to failed in a week. Wood needs to sit above roughly 20 percent moisture content for a sustained period before decay organisms get going, which is why a leak caught in one rainy season is a roofing bill and a leak ignored for three is a framing bill.
So should you foam your roof deck at all?
Often yes. Bringing the attic inside the thermal envelope is the right move in plenty of Seattle houses, especially where the ducts and the air handler live up there, or where a finished bonus room over the garage has never been comfortable. Done properly on a sound roof, these assemblies perform well and quietly for decades. That is the whole point of doing it properly.
But if your roof is near the end of its life, if there is an active leak you have been meaning to get to, or if a vented attic with air sealing and blown-in would solve your actual complaint for a third of the money, we will tell you that. Sometimes the right answer is attic insulation done the conventional way, and we would rather be right than booked.
Common questions
Does closed-cell spray foam rot a roof deck?
Not on its own. Closed-cell does not absorb and hold water, it keeps the sheathing warmer, and at proper thickness it stops indoor moisture diffusing through to the deck, so the condensation risk that gets discussed with open-cell is much lower. What closed-cell does share with every other foam is that it hides a roof leak, because the deck is no longer visible or drippy from inside. Fix the roof before you insulate under it and that risk drops to something manageable.
Is open-cell foam on a roof deck a bad idea in Seattle?
It is not automatically wrong, but it is not our default here. Open-cell is vapor open, so household moisture can reach cold sheathing and condense on it, and there are documented cases of damp OSB under open-cell roof applications. In a wet marine climate we would rather use closed-cell for this assembly and remove that variable entirely. If someone proposes open-cell on your roof deck, ask them specifically how the assembly controls vapor and how it dries.
How would I ever find a leak once the foam is on?
Mostly by paying attention rather than by looking. Check the eaves and rakes for staining, notice whether the attic smells different, look at the roof plane from the street for new dips, and ask any roofer who goes up there how the deck felt underfoot. On the outside, keep the roof maintained normally: flashing, valleys, penetrations, and gutters. You are compensating for the loss of the ceiling stain warning system with routine attention instead.
Do I need a vapor retarder with open-cell under my roof?
In colder climate zones, generally IECC zone 5 and above, codes typically require one for unvented roof assemblies with open-cell. Seattle sits in zone 4C marine, which is not in that group, so the requirement does not apply here in the same way. We treat that as context rather than a conclusion: what is required for your particular job gets confirmed with your local building department or inspector before anything is sprayed. We will do that as part of the estimate rather than leaving you to guess from a web page.
My roof is 18 years old. Should I insulate the deck now or reroof first?
Reroof first, almost every time. Foaming the underside of a deck you are going to tear the top off in a few years means paying for the insulation twice, and it means living with a hidden deck during the years the roof is least reliable. The sequence that saves money is roof, then inspect the sheathing while it is exposed, then insulate.
Can spray foam be removed from a roof deck if something goes wrong?
Yes, but it is slow, expensive, and unpleasant, and closed-cell in particular bonds hard enough that removal is mechanical and can damage the sheathing. That cost is exactly why we are careful about which product goes on a roof deck and what condition the roof is in first. The cheapest version of this job is the one that gets specified correctly at the start.