The State of Hail in America 2026

Hail is the most expensive severe-weather peril that almost nobody measures well. The federal government publishes the raw ingredients — storm reports, radar grids, damage surveys — but it publishes them as data files for meteorologists, not as answers for the people whose roofs, cars and businesses are underneath. Ask a simple question like "how often does my city get hail?" and you will find blog posts citing a decade-old top-ten list, an insurance trade group's national loss total, and almost nothing in between.

This report is our attempt to fix that. It aggregates every confirmed US hail report from the 2025 calendar year — 12,363 reports, of which 8,832 were one inch or larger — against 16,394 mapped US cities and towns, and cross-checks all of it against 22,031 radar-derived hail swath polygons covering 333 storm days. Every figure below is measured, not modeled, and the section at the end explains exactly where each one comes from and where it stops being reliable.

Use it to understand your market. If you sell roofing, insurance, restoration or anything else that moves when the sky opens, the questions that matter are where, how often, how big, and when — and all four are answerable. Read it with your service map open.

The 2025 Hail Season in One Table

The second-to-last row is worth pausing on. A six-inch hailstone fell in West Texas on May 25, 2025 — larger than a softball, and one of 63 confirmed stones of four inches or more that year. Those are the events that make the news.

The last row is the one that matters more. Radar detected damaging hail on 333 days; ground observers confirmed it on 232. Hail fell on roughly a hundred more days than anyone reported, because most hail falls where nobody is standing.

The Hail Belt: States Ranked

Ranked by confirmed reports of 1.00"+ hail during 2025.

Two columns tell different stories, and the difference is worth understanding. Reports measure intensity of activity; days measure frequency of exposure. Missouri and South Dakota make the point: Missouri logged nearly twice South Dakota's reports across fewer days, meaning Missouri's hail arrives in concentrated, heavily-witnessed outbreaks over populated ground, while South Dakota gets hit far more often across country where fewer people are watching.

Tennessee, Wisconsin and Kentucky in the top twelve will surprise anyone working from the traditional "Hail Alley" map. That map is decades old, and 2025 does not support it as an exclusive story — the eastern half of the belt is genuinely active.

One honest caveat on this table: confirmed reports require a witness, so this ranking partly measures where people live and drive. A storm over empty rangeland and an identical storm over a suburb do not generate the same number of reports. We deal with that bias directly in the methodology, and it is the reason we report confirmed counts rather than dressing up radar estimates as certainties.

The Most-Hit Cities

Cities over 100,000 people, ranked by the number of separate days a confirmed 1"+ hail report landed within ten miles.

Oklahoma City took damaging hail on eleven separate days in a single year. For a homeowner there, that is not a once-a-decade catastrophe to insure against — it is a recurring annual condition. Colorado Springs is the density outlier: 64 reports across only 8 days, meaning when it gets hit, it gets hit thoroughly and across the whole city at once.

Nearly Half the Country Sits Under It

This is the finding that reframes everything else.

Forty-five percent of the mapped US population lived in a town that had confirmed damaging hail within ten miles during 2025. Not over a decade. In one year.

And that is the floor, because it counts only hail somebody confirmed. Radar found damaging hail on 101 more days than ground observers did, which means the true exposed share is higher — we simply decline to publish a number we cannot stand behind.

The gap between hail that fell and hail that got acted on is the entire economic story of storm restoration. Most of those 122 million people never had their roof looked at, because nothing told them to. That is what a free hail damage lookup is for on the homeowner side, and what storm maps and smart canvassing are for on the contractor side. What those homeowners actually did next is the subject of The Homeowner Storm Damage Report 2026.

When Hail Happens

Distribution of confirmed 1"+ hail reports by month.

March through June accounted for 77% of the year's damaging hail, and May alone for 30%. For contractors, that compresses an entire year of demand into a window that is over before most companies have finished hiring for it — which is why the storm surge in The State of Roofing Labor 2026 is the operational problem it is. For homeowners, it means a roof inspected in September was very likely hit in April or May and has been quietly shedding granules ever since.

September's small bump above August is real and worth noting — a secondary autumn peak that catches companies who have already stood their crews down for the year.

How Big the Stones Get

Size is what separates cosmetic damage from a full replacement, and the distribution is heavily weighted toward the small end.

The practical line is 1.25 inches. Below it, damage is real but inconsistent and often disputed as wear; at or above it, functional damage is the expectation rather than the argument. That single threshold decides more claim outcomes than any other variable — the National Weather Service hail size reference is the standard both adjusters and contractors work from.

Which makes the top row the most consequential number in this report. Nearly half of all damaging hail lands in the size band where the outcome depends on documentation rather than severity. Two identical 1.1" roofs get different answers based entirely on whether somebody photographed the bruising properly — the mechanism behind the denial data in The Roofing Insurance Claim Outcomes Report 2026 and the fundamentals in what hail damage looks like on a roof.

What Radar Sees That Nobody Reports

Across the fifteen core hail-belt states, comparing radar-detected hail days against confirmed ground reports:

Radar found damaging hail on 36% more days than ground observers confirmed it, and that is inside the hail belt where spotter networks are densest and the public is most storm-aware. In thinly-populated country the gap is far wider.

This is the single best argument for looking a storm up rather than waiting to hear about it. A hailstorm that generates zero reports still damaged every roof it crossed. It simply crossed ground where nobody filed anything.

It is also why we publish confirmed counts throughout this report rather than radar counts — radar is better at coverage and worse at certainty, and a report built for citation should lead with what can be defended. How accurate are hail maps? covers the tradeoff in detail, and how to read a hail map covers what to do with one.

What the Data Means

If you own a home in the hail belt: frequency is higher than intuition suggests, and damage is quieter than you expect. Forty-five percent of the mapped population had confirmed damaging hail within ten miles in a single year, and a roof does not have to leak to be finished. Check whether hail has actually hit your address rather than waiting for a ceiling stain — our free hail history lookup is built for exactly that question, and do I have a hail damage roof claim? walks through what to do with the answer.

If you sell roofing or restoration: the constraint in this industry has never been damage supply. Over seven thousand American towns took confirmed damaging hail in 2025 and a small fraction of those roofs ever became a claim. The constraint is knowing which addresses, how big, and how confidently — before the neighborhood is already worked. That is the premise of hail maps for roofers.

If you write or price insurance: the size-band distribution above is the argument against treating hail as a single peril. Nearly half of all damaging hail sits in a band where outcomes hinge on documentation quality, which means loss ratios in that band are driven by contractor practice as much as by weather.

Methodology

Confirmed reports. Every count, ranking, monthly share and size band in this report is derived from NOAA Storm Prediction Center severe hail reports for the 2025 calendar year — 12,363 reports, of which 8,832 met the 1.00" threshold. These are observations logged by trained spotters, emergency management, law enforcement and the public, and verified by the local National Weather Service office.

Confirmed reports carry a known and unavoidable bias: they require a witness. Hail over open country is systematically under-counted relative to hail over a suburb, so state and city rankings partly reflect population and road density. We report them anyway, and say so, because the alternative is worse — see below.

City figures were produced by joining each report to the 16,394 mapped US places by geographic distance, counting a place as hit when a confirmed 1"+ report fell within ten miles of it. "Hail days" count distinct calendar dates; a day with three separate storms counts once. Population figures are place-level and are the mapped population of those places, not a household or property count — we do not hold housing-unit data and have not estimated one.

Radar figures come from NOAA MRMS-derived hail swath polygons — 22,031 polygons at the 1.00" threshold across 333 days in 2025. Cumulative swath area sums each day's polygons, so ground hit on multiple days is counted multiple times; it is a measure of storm activity, not of unique land area.

Why radar is used only for coverage comparisons. We validated radar-detected hail days against confirmed reports state by state. Across the core hail-belt states the two track each other at a stable and physically sensible ratio of roughly 1.5 to 2.3 radar days per confirmed-report day. Outside that region the ratio breaks down badly — coastal, Canadian-border and mountainous-West states run ratios from 3 to 16, which reflects radar beam artifacts over water, at extreme range and against terrain, not real hail. We therefore do not publish radar-derived state or city rankings, and the radar comparison in this report is scoped to the validated hail-belt states only. Any figure here that would have depended on unvalidated radar has been left out rather than estimated.

Not included. We hold no insured-loss or damage-cost data and publish none; the Insurance Information Institute's hail statistics and NOAA's billion-dollar disaster record are the appropriate sources for that. Alaska, Hawaii and US territories are excluded. All figures cover the 2025 calendar year and were generated on 2026-08-21; the SPC occasionally adds late-verified reports, so counts may drift marginally over time.

Related Reading

• How to Read a Hail Map — what the colors and confidence bands on a hail map actually mean.
• How Accurate Are Hail Maps? — radar estimates versus ground truth, and where each one fails.
• The Homeowner Storm Damage Report 2026 — what the affected homeowners actually did next.

This report is for informational purposes only.