What Size Hail Damages a Roof? One Inch for Some Roofs, Two for Others.

Ask ten roofers what size hail damages a roof and eight will say "an inch." It's a decent rule for one kind of roof. For the rest it's either too generous or too strict, and on a storm list that difference decides which half of the street you knock first.

The best public data on this comes from Haag Engineering in Dallas, whose engineers have been firing ice balls at roofing since 1963. In one widely cited study presented to the American Meteorological Society, they launched solid ice stones at nine common roof coverings, ten shots per size, at the speed a hailstone of that size actually falls. Then they counted how many shots did real damage: a fractured mat, a puncture, or granules knocked off to expose the asphalt.

Here's what they found, as the share of impacts that did damage.

Two things jump out. An 11-year-old 3-tab fiberglass roof goes from untouched at 1 inch to damaged on most hits at 1.25 inches. And an 11-year-old laminated roof didn't take a single damaging hit until 1.5 inches. That's the shingle most Front Range and North Texas subdivisions built after about 2005 are wearing.

So "an inch" is really "an inch and a quarter for old 3-tab, an inch and a half for architectural, two inches for S-tile."

Why a quarter-inch matters so much

Hailstone energy doesn't grow in a straight line. A bigger stone is heavier and falls faster, so the energy stacks. Using the terminal velocities in the same study:

Going from 1 inch to 1.5 inches multiplies the energy by more than eight. Going from 1.5 to 2 inches nearly triples it again. That's why a laminated roof can shrug off 1.25-inch hail and be ruined by 1.75-inch hail from the next cell over. The stone didn't get a little bigger. It got several times harder-hitting.

For a roofer that means the size line on your list isn't a detail. Setting it a quarter-inch too low sends reps to hundreds of roofs that were never hurt.

The lab is the worst case, and that cuts both ways

Before you treat the chart as gospel, two caveats the engineers flagged themselves.

Lab ice is harder than hail. Freezer ice is denser than natural hailstones, which are layered with rime and air pockets. And every lab shot hit dead-on, at 90 degrees. Real hail is often softer and glances off a pitched roof. On a calm night, the real-world threshold can run a little higher than the table.

Wind pushes the other way. Wind-driven hail hits the windward slope harder and at a flatter angle. A 1.25-inch stone in a 50 mph gust carries more energy than the chart's still-air number, which is why you'll see one side of a house shredded and the other side fine. We covered what wind alone does to a roof in the wind damage post.

Age matters most of all. Every shingle in that table was about 11 years old. A 20-year-old 3-tab that's lost its oils is more brittle still, and a 3-year-old laminated shingle is more flexible. That's the core argument of our roof-age post: the same storm is a claim at one address and nothing at the neighbor's.

What about Class 4?

A Class 4 rating (UL 2218) means a sample survived a 2-inch steel ball dropped from 20 feet without cracking. It's a real step up. It is not hail-proof. Class 4 shingles still lose granules and scuff under large stones. The damage is just less likely to crack the mat.

The bigger practical issue on a Class 4 or metal roof is paperwork, not physics. Many Colorado and Texas policies with impact-resistant discounts also carry a cosmetic damage exclusion, so dents and scuffs that don't let water in often aren't covered. We walked through how that endorsement works in the cosmetic damage post. For list-building, treat Class 4 and metal roofs as a separate, low-priority bucket unless the stones were well past 2 inches.

Don't compare the chart to the biggest number on a hail map

Here's where roofers get burned. The chart is about the stone that hit the roof. Most hail maps show something else: the biggest radar-estimated stone anywhere near the address.

That number runs high. When we checked our own map against 785 Colorado spotter reports, radar found 99% of confirmed hail events. But the "max near this house" figure read about 1.25 inches larger than what spotters measured on the ground.

Put those two facts together and the problem is obvious. A map shows "2.25 inches" near a 2010 laminated roof. You read it against the chart and see near-certain damage. The stone that actually landed on that roof may have been closer to an inch, which the chart says a laminated shingle usually survives. That's an inspection that goes nowhere and an afternoon gone.

The fix is to use the size measured at or right next to the address, not the max in a wide circle, and to lean on ground reports where they exist.

Setting your own cut line

Here's how we'd turn the chart into a knock order, using radar-estimated size at the address plus roof age from county records:

1. Top of the list: 3-tab or unknown shingle, roof 15+ years old, 1.25"+ at the address. The chart says most hits do damage, and on a brittle roof that age the bruising is plain to see on inspection.
2. Second pass: laminated, 12+ years old, 1.5"+ at the address. This is the bulk of suburban Colorado and DFW, and where most of the money is.
3. Third pass: laminated under 10 years old, 1.75"+. Real claims live here, but expect more denials and plan for it.
4. Skip for storm work: S-tile under 2", Class 4 and metal unless the stones were far past 2", and any roof replaced after the storm date.

Run the arithmetic on your own crew. If a rep burns 45 minutes on an inspection that goes nowhere, and a quarter-inch too-low cut line adds 30 of those a week across the team, that's more than 22 hours a week of paid time on roofs the lab data says were fine. Plug your own rep cost and close rate into the cost-per-job calculator and you'll see how quickly a sloppy size line shows up in cost per signed job.

The short version

• Old 3-tab: starts breaking around 1 to 1.25 inches.
• Laminated / architectural: around 1.5 inches; reliably by 1.75.
• Wood shingles and shakes: 1.25 to 1.5 inches.
• Concrete tile: flat tile varies; S-tile holds out to about 2 inches.
• Class 4 and metal: damage is possible above 2 inches, but a cosmetic exclusion often decides the claim.

Radar tells you where hail fell. Roof type and roof age tell you whether it mattered. Our territories combine both: every home inside the radar footprint, scored on radar-estimated stone size at the address and roof age, for one roofer per city. See what's been radar-estimated near you on the Colorado hail map, and if your city is still open, check the territory map. First come.