County and state lookup
Fatal crash rate by county
Type a county and get its fatal crash rate: the crashes NHTSA recorded there, divided by the people who live there, per 100,000. Every input is on the page, and so is what the record cannot say.
No county matched that text. Pick one from the list as you type, or write it as “Harris County, Texas”.
Or open a state below for its rate per mile driven and its county list.
A worked example: Harris County, Texas
Harris County, Texas2024 · example
11.0fatal crashes per 100,000 residents in 2024
In 2024, NHTSA recorded 551 fatal crashes in Harris County, where 4,996,331 people lived. 583 people died in them. Over 2020 to 2024 the average was 10.7.
The road is a scale of fatal crashes per 100,000 residents, 0 to 40. Every county and state page draws its own.
- Formula
- fatal crashes ÷ residents × 100,000
- 2024
- 551 ÷ 4,996,331 × 100,000 = 11.0
- United States
- 36,297 ÷ 340,003,797 × 100,000 = 10.7
Every state
Fatal crashes per 100,000 residents in 2024, alphabetical. State pages add the rate per 100 million miles driven.
- Alabama 17.3
- Alaska 8.6
- Arizona 14.8
- Arkansas 17.7
- California 9.1
- Colorado 10.7
- Connecticut 7.8
- Delaware 11.5
- District of Columbia 6.7
- Florida 12.6
- Georgia 11.7
- Hawaii 6.8
- Idaho 10.9
- Illinois 8.5
- Indiana 11.3
- Iowa 10.0
- Kansas 10.6
- Kentucky 14.5
- Louisiana 15.3
- Maine 11.9
- Maryland 8.8
- Massachusetts 4.9
- Michigan 10.0
- Minnesota 7.4
- Mississippi 23.0
- Missouri 14.1
- Montana 17.0
- Nebraska 11.1
- Nevada 11.6
- New Hampshire 8.5
- New Jersey 6.7
- New Mexico 17.8
- New York 5.2
- North Carolina 13.7
- North Dakota 10.6
- Ohio 9.1
- Oklahoma 14.5
- Oregon 11.5
- Pennsylvania 8.1
- Rhode Island 4.3
- South Carolina 17.3
- South Dakota 14.5
- Tennessee 15.1
- Texas 12.1
- Utah 7.2
- Vermont 8.2
- Virginia 9.8
- Washington 8.5
- West Virginia 13.5
- Wisconsin 8.9
- Wyoming 17.4
How the rate is built
- Numerator: fatal crashes from NHTSA’s Fatality Analysis Reporting System, yearly national CSV files 2020 to 2024 (2024 is the Annual Report File until the final file replaces it). One row per crash; deaths are the
FATALSfield; the county is the FARS county code joined to the Census FIPS code. - Denominator: the Census Bureau’s July 1 population estimate for the same year, Vintage 2025 county totals. State pages add FHWA’s table VM-2 vehicle-miles for a rate per 100 million miles, the form in 23 CFR 490.207; FHWA publishes no county vehicle-miles.
- Five-year average: total crashes ÷ total population over 2020 to 2024 × 100,000.
- Gaps, stated on the page they affect: Connecticut’s legacy counties (counts, no rate); crashes coded to an unknown county (state only, about 0.01%); Valdez-Cordova, Alaska, split in 2019 with no successor (state only).
- Not here: rankings, grades, injury crashes, congestion, or the risk to any one driver or trip.
Informational only, not professional advice. Last reviewed: October 2026. Snapshot built 2026-10-28.
Frequently asked questions
What counts as a fatal crash?
A crash on a public road in which at least one person died within 30 days, as recorded in NHTSA's Fatality Analysis Reporting System (FARS). One crash can have several deaths; the rate here counts crashes, and the deaths are shown beside it.
Why is there no rate for my Connecticut county?
NHTSA still codes crashes to Connecticut's eight legacy counties, but the Census Bureau has published populations only for the nine planning regions that replaced them since 2022. Those county pages show counts; the Connecticut state page has both rates.
Why is a small county's rate so high (or zero)?
With 5,000 residents, one fatal crash is 20 per 100,000 and none is zero. Every page prints the population and the five-year average so a one-year rate is never read alone. The site does not rank counties for the same reason.
Does this cover injuries, all accidents or traffic?
No. FARS is fatal crashes only. Injury and property-damage crashes sit in fifty separate state files and congestion data is proprietary, so no page here says anything about them.