Every Age Group Got Safer in 2025. If You're Over 65, None of It Applies to You.

NHTSA just dropped its 2025 sub-category fatality estimates, and they read like a victory parade. Total traffic deaths fell 6.7 percent to 36,640, the fatality rate hit 1.10 per 100 million vehicle miles traveled — the second-lowest full-year figure in American history — and children under 15 saw a 12 percent drop while adults 25 to 34 improved a full 13 percent.^[1] Speeding deaths cratered 11 percent, ejection fatalities plummeted 13 percent, and it is, by almost every measure, a genuinely good year.

Almost.

People 65 and older accounted for 8,039 traffic deaths in 2025, down from 8,060 in 2024, a 0.3 percent change that NHTSA rounds, correctly, to zero.^[1] Twenty-one fewer deaths. Adults 25 to 34 saved 891 lives in the same period, an improvement ratio of 42 to 1.

Run the counterfactual: if the 65-and-older cohort had matched the national 6.7 percent decline, 2025 would have recorded roughly 7,520 senior traffic deaths instead of 8,039, meaning 519 excess deaths occurred because whatever made everyone else safer did not reach people past retirement age. In 2024 seniors accounted for 20.5 percent of all traffic fatalities; in 2025 that share climbed to 21.9 percent, and by November and December it hit 24 percent.^[1] Hold that trajectory and seniors will account for one in four American road deaths within two years.

So why didn't any of the improvements reach them? Newer vehicles saw a 12 percent occupant fatality decline, but 63 percent of fatally injured passenger vehicle occupants were in vehicles ten years old or older.^[1] Unrestrained occupant deaths fell 13 percent, but seniors already buckle up at rates above the national average. Speeding deaths dropped 11 percent, but seniors are underrepresented among speeders. Every mechanism that drove 2025's improvement targeted behaviors and exposures that skew young, leaving the oldest cohort with nothing.

Then there is the variable that no safety system addresses: fragility. A 35-year-old and a 75-year-old in the same model year Camry, wearing the same seatbelt, hit at the same speed, face profoundly different survival odds because chest wall compliance, cervical spine tolerance, and aortic tear threshold all decline with age in ways that no crash test dummy is calibrated to capture. The Hybrid III 50th-percentile male used in federal crash testing is perpetually 35, with no osteoporosis and no pacemaker, and will remain so until regulators decide he shouldn't be.^[2] IIHS data confirms that per-mile fatal crash rates climb sharply past 75, driven by increased injury severity rather than increased crash frequency: seniors don't crash more often, they just die more often when they do.^[3]

Strongest counterargument: the flat count may reflect slight per-capita improvement, because the 65-plus population grew between 2024 and 2025. Fair point, but "your death rate improved by a fraction while your children's dropped by double digits" is actuarial-footnote comfort, not a safety win.

Limitations: These are ratio-adjusted projections from incomplete FARS data, not final counts; alcohol impairment data is excluded due to reporting lag; and senior VMT-by-age data is not published annually, making age-specific rate calculations approximate.

If you have a parent or grandparent driving a vehicle more than ten years old, check their car's crash test rating against current standards, not the year it was built, and look up their VIN at nhtsa.gov/recalls. Understand that a five-star rating earned with a crash test dummy built to the tolerances of a 35-year-old tells you how well the car protects a 35-year-old.