Every Age Group Got Safer on America’s Roads Last Year. Except the Biggest One.
NHTSA released its full subcategory breakdown of 2025 traffic fatalities this month, and the headline number looks genuinely encouraging: 36,640 dead, down 6.7 percent from 2024, with a fatality rate of 1.10 per 100 million VMT, the second-lowest ever recorded. Speeding deaths fell 11 percent, unrestrained occupant deaths dropped 13 percent, and even pedestrian fatalities declined 8 percent.[1]
Then you get to Table 1, Row 65+, and the number sitting there is 8,039 dead, essentially unchanged from 8,060 in 2024. Twenty-one fewer seniors in a year when 2,614 fewer people died overall. Adults 25 to 34 improved 13 percent, children under 15 improved 12 percent, drivers 35 to 44 shed 9 percent, even the 55-to-64 bracket managed a 3 percent drop. But the 65-and-over cohort, now the single largest age group in the fatality tables and the fastest-growing demographic in the country, stayed perfectly, stubbornly flat. As a share of all traffic deaths, they climbed from 20.5 percent to 21.9 percent, trending toward one in four by year’s end.
Run a simple counterfactual and the gap gets a body count attached to it. If 65+ had improved at the national average of 6.7 percent, roughly 7,520 would have died instead of 8,039, a difference of 519 excess deaths attributable to one cohort’s failure to participate in a broad national improvement. Benchmark against the best-performing group and the gap widens past a thousand.
Physiology is part of the explanation, and the honest part at that. A crash that bruises a 35-year-old’s sternum fractures a 75-year-old’s ribcage, and aortic tears and subdural hematomas escalate relentlessly with age. But fragility explains the high baseline; it does not explain zero improvement while everyone around you gets safer. FARS per-model data suggests a vehicle-fleet story underneath the demographic one: the cars historically associated with elderly drivers carry death rates from another era, with the Chevrolet Impala at 5.0 per 100 million VMT, the Buick LeSabre at 2.67, and the Mercury Grand Marquis at 2.29, compared with ADAS-equipped vehicles like the Mazda CX-5 at 0.12 or the Kia Telluride at 0.04.[2] Between a LeSabre and a Telluride, that ratio is 67 to 1, and NHTSA’s own data shows deaths in newer vehicles fell 12 percent last year while deaths in older vehicles fell only 9 percent.
Demographic math offers the strongest counterargument: Baby Boomers are swelling the 65+ population faster than any other age group, meaning more exposure miles even at constant risk, and FARS aggregate data does not break out per-model fatality rates by driver age, making the vehicle-fleet cross-reference suggestive rather than conclusive. Both are real confounds. But population growth alone does not produce a flatline at exactly zero improvement while safety technology hauls every other cohort downward by 3 to 13 percent.
IIHS awarded Top Safety Pick+ to 45 vehicles for 2026, thirteen priced under $30,000, and AEB is standard on more than 95 percent of new vehicles sold today.[3] If someone you love is over 65 and still driving a vehicle built before 2015, every year they keep it is a year they sit outside a national safety improvement saving thousands of lives around them. Check their VIN at nhtsa.gov/recalls, look at the IIHS list for something with AEB and solid structural ratings, and have the conversation before the data has to.
Sources & References
- NHTSA, Early Estimates of Motor Vehicle Traffic Fatalities and Fatality Rate by Sub-Categories in 2025, DOT HS 813 829, July 2026. crashstats.nhtsa.dot.gov
- NHTSA, Fatality Analysis Reporting System (FARS), 2014–2023. nhtsa.gov
- IIHS, 2026 Top Safety Pick and Top Safety Pick+ Awards. iihs.org
Source: NHTSA FARS 2014–2023; NHTSA Early Estimates sub-categories report, DOT HS 813 829 (July 2026). “Excess deaths” is a simple counterfactual, not a causal estimate. FARS per-model rates are aggregated across all driver ages. See methodology for caveats.