Fatal Crashes per 100 Million Vehicle-Miles, Explained

The short answer

A fatal crash rate per 100 million vehicle-miles divides the fatal crashes recorded in a state by the miles driven there, then scales the result to 100 million miles. It answers a different question from the per-person rate: not how many fatal crashes per resident, but how many per mile of driving. A state with long distances and heavy through-traffic can have a high per-person rate and an ordinary per-mile rate.

fatal crashes ÷ vehicle-miles × 100,000,000 = fatal crashes per 100 million miles

The crashes come from NHTSA’s FARS file. The miles come from the Federal Highway Administration’s Highway Statistics table VM-2, which estimates annual vehicle-miles of travel for each state from traffic counts. Federal highway safety rules, in 23 CFR 490.207, define the fatality rate in exactly this per-100-million-miles form.

A worked example

In 2024, FARS recorded 3,774 fatal crashes in Texas. FHWA estimates 308,461 million vehicle-miles were driven there that year.

3,774 ÷ 308,461,000,000 × 100,000,000 = 1.22

So Texas had 1.22 fatal crashes per 100 million miles driven. The United States as a whole was 1.10 (36,297 crashes over 3,294,031 million miles).

The two rates side by side

The per-person rate uses residents as the denominator; the per-mile rate uses driving. Where people drive more miles per head, the per-person rate runs higher than the per-mile rate would suggest, because each resident is exposed to more road. Neither is more correct. They measure different things, which is why state pages show both with the inputs beside them.

Why county pages have no per-mile rate

FHWA publishes vehicle-miles by state, not by county. Some state transportation departments estimate county miles, but there is no national file built the same way everywhere, and a rate built on fifty different methods would not be comparable. So county pages show the per-person rate only, and say so. The state a county belongs to is one click away for the per-mile figure.

What the per-mile rate leaves out

• Who was driving. Vehicle-miles are an estimate of total travel on a state’s roads, from residents and visitors alike. The rate says nothing about any one driver.
• Which roads. Miles on an interstate and miles on a rural two-lane road are added together. The state page lists how many fatal crashes FARS coded to each road type, which is a different question.
• Timing of the files. FARS and VM-2 are published on different schedules. The page names the year used for each; if a year has crashes but no miles yet, the per-mile rate uses the latest year that has both.
• Estimation. Vehicle-miles are modelled from traffic counts, not counted one by one. The crash count is a count; the mileage is an estimate.

What this guide does not cover

It does not cover the per-person rate in detail, which has its own guide, or how FHWA builds its mileage estimates. It does not say whether any state’s rate is high or low, and it does not rank states.

Look up a state

Every state page shows both rates for 2020 to 2024, with the crashes, the population, the vehicle-miles and the year of each file.

This guide is for informational purposes only and is not professional advice. A fatal crash rate describes fatal crashes recorded in one place over one year. It is not all crashes, not injuries, and not the risk to any one driver, road or trip.

Last reviewed: October 2026 · Against the FARS 2024 Annual Report File and FHWA Highway Statistics table VM-2 for 2024.