How FMCSA Crash Reporting Works — and Its Known Gaps
A carrier's crash history is one of the more intuitive safety signals available — it's easy to understand what a crash count means in a way a calculated BASIC measure sometimes isn't. But like any dataset, FMCSA's crash reporting has a specific scope and specific limitations, and understanding both makes the number more useful, not less.
Where the data comes from
Crash reports reach FMCSA's system through state agencies, which submit qualifying crash records involving commercial motor vehicles. A crash generally qualifies for reporting based on severity thresholds — a fatality, an injury requiring transport for medical treatment, or a vehicle disabled and towed from the scene. Because the trigger is a state-reported crash, the timeliness and completeness of the data depends partly on how quickly and thoroughly each state processes and forwards its records.
What's captured for each crash
A crash record includes the date, location, and severity outcome (whether there were fatalities, injuries, or a tow-away), along with details like weather conditions, road surface, lighting, and the vehicle configuration involved. This gives useful context beyond a bare count — a pattern of crashes in poor weather conditions reads differently than a pattern in clear conditions with no obvious contributing factor.
What crash data doesn't tell you
Critically, a crash being reported doesn't mean the carrier was at fault. FMCSA's crash reporting captures that a commercial vehicle registered to a carrier was involved in a qualifying crash — it does not include a fault determination. A carrier that was struck by another driver running a red light shows the same "crash reported" status in this dataset as one that caused a preventable accident through its own negligence. This is arguably the single most important caveat to keep in mind when reading crash history: the count reflects involvement, not blame.
Known gaps and limitations
Because reporting relies on state submission, there's inherent variability in how consistently and quickly different states report qualifying crashes, which can affect comparisons between carriers based in different states. There can also be a reporting lag between when a crash occurs and when it appears in the federal dataset, meaning very recent crashes may not yet be reflected. And because reporting is based on severity thresholds, minor incidents that don't meet those thresholds — a low-speed parking lot fender-bender, for instance — generally won't appear at all, even though they happened.
Why crash rate, not just crash count, matters
A raw crash count means very different things for a two-truck operation and a two-hundred- truck fleet. That's why CarrierPedia expresses crash activity as a rate — crashes per 100 power units — alongside the raw count, and shows how that rate compares to peers in the same state and cargo category. See What an Out-of-Service Rate Says About a Carrier for the same logic applied to inspection outcomes.
Reading crash history responsibly
Given the fault-neutral nature of the data, the most defensible way to use crash history is as one input among several — alongside out-of-service rates, BASIC scores, and how long a carrier has been operating — rather than a standalone verdict on a carrier's safety. See How to Verify a Carrier Before Booking a Load for how it fits into a fuller check, or look up a specific carrier's crash record with DOT Number Lookup.