Road safety, trucking and the law, explained.
Statistics, datasets and original analysis, with every figure traced to its source.
Each national crash number comes from a system built to answer one question. A guide to the census, the sample, the carrier files, the one-time causation studies and the compilations that repackage them.
Editorial Desk7 min read
Texas police reported 38,533 crashes involving a commercial motor vehicle in 2025, and 528 people died in them. Here are the yearly and county counts, where they come from, and what each one measures.
Editorial Desk7 min read
| Dataset | Publisher | What it counts | Watch for |
|---|---|---|---|
| FARS | NHTSA | Every crash on a public road in which someone died within 30 days. Annual census since 1975. | Fatal crashes only. Final files lag the calendar year. |
| CRSS | NHTSA | A nationally representative sample of police-reported crashes. Replaced GES starting with 2016 data. | Estimates with margins, not counts. Crashes never reported to police are missing. |
| MCMIS | FMCSA | Truck and bus crashes and roadside inspections reported by the states. | Completeness depends on each state’s reporting. |
| TxDOT CRIS | Texas DOT | Reportable crashes in Texas, from officer crash reports. | State definitions differ from federal ones. |
| LTCCS | FMCSA & NHTSA | One-time study of 967 injury and fatal large-truck crashes, 2001 to 2003. | Associated factors are not causes. The data are two decades old. |
| NMVCCS | NHTSA | On-scene causation survey of light-vehicle crashes, 2005 to 2007. | The “94 percent” figure comes from here and is often misquoted. |
We start from the original file or report, not a press release or a second-hand compilation. Each figure on a data page carries the dataset, the year and, where it matters, the table it came from. When a number needs context to be read correctly, such as a sample estimate with a margin of error or an associated factor that is not a cause, the context sits next to the number rather than in a footnote.
When we cannot verify a figure against an original source, we do not publish it. Draft data pages mark those gaps openly until the figure is confirmed. Corrections to published figures follow our corrections policy and are noted on the page.
Requests for underlying tables, methodology questions and proposed research collaborations go to editorial@bttlaw.com.
