A Class A roof improves a home’s odds of surviving a wildfire, but post-fire damage-inspection data shows the roof covering alone is not what separates a house that stands from one that burns. Across CAL FIRE’s Damage Inspection Program (DINS) records and Insurance Institute for Business and Home Safety (IBHS) field studies, roof construction ranks as a secondary predictor of survival, behind how far apart homes sit and what covers the exterior walls. The largest survival gains show up only when a Class A roof is paired with ember-resistant vents, enclosed eaves, and a noncombustible zone within the first five feet of the wall.
Do Class A roofs actually save homes in wildfires?
Class A roofs help, and the empirical case for them is real, but they are a necessary part of a system rather than a standalone guarantee. In the 2025 Palisades and Eaton fires, IBHS found that homes with four hardening features together, a Class A roof plus noncombustible siding, double-pane windows, and enclosed eaves, had a 54% likelihood of avoiding damage, compared with just 36% for homes that took only a single action (IBHS, 2025). The roof matters, but the four-feature package roughly halved the damage rate versus one action alone.
The headline finding across the data is consistent: a Class A roof covering removes one of the most fire-prone surfaces on a house, yet homes with Class A roofs still burn when embers enter through vents, lodge in gutters, or ignite vegetation and combustible materials touching the wall. The roof rating is one input into survival, not the whole equation.
What “Class A” actually means: the ASTM E108 test
Roof fire classes come from ASTM E108 (equivalent to UL 790), a test that rates roof coverings against fire coming from outside the building. Class A withstands severe exposure, Class B moderate, and Class C light exposure (ASTM E108 / UL 790). The relevant part for wildfire is the burning brand test, which places a burning Douglas fir block on the roof to simulate wind-driven embers, the exact threat that ignites most homes in a wildfire.
| Roof fire class | Test exposure (ASTM E108 / UL 790) | Typical roof coverings |
|---|---|---|
| Class A | Severe; withstands a large burning brand simulating embers | Concrete and clay tile, metal, fiberglass-mat asphalt shingles, slate, most standing-seam |
| Class B | Moderate exposure | Some pressure-treated wood shakes and specific rated assemblies |
| Class C | Light exposure | Untreated wood shingles, some organic-mat products |
| Unrated | None | Aged or bare wood shake, some older membranes |
Most modern asphalt shingles carry a Class A rating when installed as a tested assembly, which is why asphalt composition covers about 78.6% of assessed U.S. buildings while wood roofs sit at 5.8% (Headwaters Economics). The wood-roof exception is where the historical damage concentrated. Our guide to Class A fire-resistant roofing materials breaks down which coverings earn the rating and how, and the full types of roofing materials by slope and form.
What post-fire data shows about roof class and survival
The strongest evidence comes from CAL FIRE DINS, which documents every structure inside or within 300 feet of a wildfire perimeter in California since 2013. A 2025 analysis in Nature Communications used roughly 47,000 structures from five major fires (2017 Tubbs and Thomas, 2018 Camp, 2019 Kincade, 2020 Glass), drawn from a larger DINS set of about 90,000 structures spanning 2013 to 2022. A machine-learning model predicted survival with 82% accuracy and ranked which features mattered most (Nature Communications, 2025).
Roof construction landed inside the top predictors but below structure separation distance, exterior siding, and year built. The table below ranks building features by their measured or observed importance for structure survival, synthesized across the DINS analysis, IBHS field work, and San Diego County structure studies. It is a relative ranking, not a single-source odds ratio.
| Building feature | Relative importance for survival | What the data shows | Primary source |
|---|---|---|---|
| Structure separation distance | Highest | Homes spaced farther from neighbors survived more often; the top model predictor | Nature Communications, 2025 |
| Exterior siding (noncombustible) | High | Wall material was among the strongest destruction predictors | Nature Communications, 2025 |
| Year built (code and materials proxy) | High | Newer homes built to WUI codes survived at higher rates | Nature Communications, 2025 |
| Enclosed eaves | High (local scale) | Among the two most important building features against loss | Syphard, Brennan and Keeley, 2019 |
| Multi-pane windows | High (local scale) | Window preparation especially important close to the home | Syphard, Brennan and Keeley, 2019 |
| Near-home zone (0 to 5 ft) | High | Over 25% vegetation in Zone 0 pushed damage or destruction risk near 90% | IBHS, 2025 |
| Class A roof covering | Secondary / moderate | In the top 10 predictors, below spacing, siding, and year built | Nature Communications, 2025 |
| Ember-resistant vents | Moderate | Vents are a primary ember entry path into attics | IBHS, 2025 |
The San Diego County work behind two of those rows mapped 1,000 destroyed and 1,000 surviving structures from fires between 2001 and 2010, finding that enclosed eaves and multi-pane windows were the most important building factors, and that defensible space adjacent to the home was as important as construction itself (Syphard, Brennan and Keeley, 2019).
Why homes with Class A roofs still burn
Embers, not direct flame contact, are the primary cause of home ignition in wildfires, and embers exploit the connections around a roof rather than the roof surface itself (IBHS). A Class A roof covering resists a burning brand, but wind-driven embers still collect in gutters, slip under roof edges, enter through unscreened attic and eave vents, and ignite debris or vegetation against the wall. The roof passes its test while the assembly around it fails.
- Vents: Embers drawn into attic vents can ignite a home from the inside, which is why ember-resistant vent screens are a separate mitigation from the roof covering (IBHS, 2025).
- Gutters and roof debris: Dry leaves in gutters and roof valleys give embers a fuel bed directly on the roof edge, regardless of the covering’s class.
- Eaves and edges: Open eaves trap heat and embers under the roofline; enclosed eaves ranked among the strongest survival features in the San Diego data (Syphard, Brennan and Keeley, 2019).
- The first five feet: Combustible mulch, fences, or plants touching the wall carry fire up to the roof from below.
How much difference does hardening add on top of a Class A roof?
The survival curve steepens sharply when a Class A roof is combined with defensible space and other hardening. In the DINS-based model, home hardening alone left only about 25% of structures surviving, hardening plus a noncombustible Zone 0 raised survival to 40%, and hardening plus Zones 0 and 1 reached 48%, a 52% reduction in losses versus no mitigation (Nature Communications, 2025). The pattern matches the IBHS field finding that stacking features, not adding one, drives outcomes.
| Mitigation package | Home outcome | Source |
|---|---|---|
| Single hardening action | 36% avoided damage | IBHS Palisades / Eaton, 2025 |
| Four features together (Class A roof, noncombustible siding, double-pane windows, enclosed eaves) | 54% avoided damage | IBHS Palisades / Eaton, 2025 |
| Home hardening alone (model) | ~25% survived | Nature Communications, 2025 |
| Hardening plus Zone 0 defensible space | ~40% survived | Nature Communications, 2025 |
| Hardening plus Zones 0 and 1 | ~48% survived (52% fewer losses) | Nature Communications, 2025 |
| Zone 0 with over 25% vegetation | ~90% risk of damage or destruction | IBHS, 2025 |
The wood-roof exception, where roof class was decisive
The clearest case that roof class alone can change outcomes is the wood shake roof. The 1991 Oakland Hills firestorm destroyed more than 3,500 homes, with wood-shingle roofs cited as a leading contributor to how fast fire spread from house to house (NFPA / Cal OES). Roughly 1.2 million U.S. buildings still carry wood shake or shingle roofs, and nearly 1 million of those sit in medium-to-very-high wildfire risk areas (Headwaters Economics). Replacing those roofs with Class A coverings would cost at least $6 billion nationally, or about $5,860 (27%) added to a typical roof job (Headwaters Economics). Where the baseline is combustible wood, upgrading to Class A produces a large, measurable survival gain. Where the baseline is already a Class A asphalt or tile roof, the marginal gain from the covering shrinks and the surrounding details take over.
What this means for homeowners
A Class A roof is a sound floor, not a finish line. Homeowners in wildfire-prone areas should treat the roof covering as the first of several linked steps rather than the single fix that buys survival. The data supports a sequence: verify the roof is a rated Class A assembly, then close the ember paths that a Class A rating does not cover.
- Confirm the roof is a tested Class A assembly, not just a Class A shingle on an unrated deck.
- Keep gutters, valleys, and the roof surface clear of leaf and needle debris.
- Install ember-resistant (1/8-inch metal mesh or listed) vents and enclose open eaves.
- Clear the first five feet around the wall of combustible mulch, plants, and fences.
- Increase spacing and reduce connective fuels between structures where possible.
Requirements vary by state and jurisdiction, tracked in our companion report on state-by-state Class A roof mandates, and insurers increasingly price roof class into coverage decisions, as covered in which roofs insurers reward. Outcomes depend on fire intensity, wind, and neighborhood density, so no single feature guarantees survival.
Methodology
This report synthesizes post-fire structure-outcome data rather than code language. The core dataset is CAL FIRE DINS, which records structures damaged, destroyed, or surviving within 300 feet of California wildfire perimeters since 2013. Feature-importance and survival percentages come from a 2025 Nature Communications analysis of about 47,000 DINS structures across five fires, an IBHS field assessment of more than 250 properties in the 2025 Palisades and Eaton fires combined with DINS records on more than 30,000 structures, and a peer-reviewed San Diego County study of 2,000 matched structures (Syphard, Brennan and Keeley). Roof-class definitions follow ASTM E108 / UL 790. Building-stock and cost figures come from Headwaters Economics. Every statistic is attributed inline to a named source and year.
Limitations
These findings are associational, not a controlled experiment, so feature importance reflects correlation within observed fires, not a guaranteed causal effect for any single home. Roof construction’s ranking is partly obscured because year built correlates with both roof type and building code, so the roof’s independent effect may be understated in the models. Most data is drawn from California WUI fires and may not transfer directly to other regions, fuel types, or extreme wind events. DINS captures roof material broadly and does not always distinguish a tested Class A assembly from a Class A covering on an unrated deck. Percentages are rounded as reported by each source.
Reviewed by The Roofing Brief Team. Last reviewed July 2026.