For Every Death on American Roads, 62 People Are Injured. That Ratio Is Climbing.

NHTSA dropped its 2024 final count on April 1: 39,254 killed, 2.42 million injured.^[1] Early 2025 estimates peg fatalities around 36,640, a 6.7% decline that NSC's monthly data corroborates.^[2] Headlines celebrated. Nobody mentioned the injuries.

That ratio used to be lower. In 2019, NHTSA recorded 36,096 deaths and 2.74 million injured, a ratio of 75.9:1. In 2021, 42,939 died against 2.50 million injuries: 58.2:1.^[3] The pandemic spike killed more people per crash. As fatalities recede, the ratio climbs back. But injuries remain anchored above 2.4 million regardless of which direction deaths move. That number refuses to drop.

Two forces explain this.

Crashworthiness Is Winning. Crash Avoidance Isn't.

Modern vehicles are exceptional at keeping occupants alive during impacts. FARS data from 2014 to 2023 shows the gap by class. In fatal crashes involving an SUV, the SUV occupant dies 52.3% of the time. For sedans, it's 64.3%. Pickups: 48.5%.^[4] As Americans migrate from sedans to crossovers and SUVs, the fleet-weighted lethality per crash drops mechanically. Sedan market share fell from roughly 30% of new sales in 2018 to about 17% in 2024.^[5] Replace a million sedan-class crashes with SUV-class crashes, and occupant deaths decline 18.7% purely from the math: (0.643 − 0.523) / 0.643 = 0.187.

Injuries don't get the same benefit. A crash that would have killed a Cavalier driver at 45 mph merely hospitalizes an Equinox driver at the same speed. One fewer death. One more injury. Survivability improvement, not prevention.

The Mileage Trick

NSC reports the December 2025 mileage death rate at 1.16 per 100 million vehicle miles, down from 1.30 a year earlier.^[2] But total miles driven rose 2.7% from 2023. Americans are driving more. The per-mile rate improvement looks dramatic, but the absolute injury count suggests those extra miles generate just as many crashes. Better crumple zones absorb the energy. ERs absorb the patients.

What Would Actually Move the Injury Number

Crashworthiness saves lives inside the vehicle. Crash avoidance prevents the event entirely. IIHS research shows automatic emergency braking cuts front-to-rear crashes by 50% and pedestrian strikes by 27%.^[6] The 2026 federal AEB mandate requires all new light vehicles to include the technology by 2029. That timeline leaves 3 more model years of new vehicles without mandatory AEB, plus a fleet turnover lag of 12 to 15 years before penetration reaches critical mass.

If you drive a vehicle built before 2020, odds are it lacks pedestrian AEB. If it was built before 2018, it may lack AEB entirely. Check whether your specific model has it. If not, aftermarket forward-collision warning systems exist from Mobileye and others for $800 to $1,200 installed. They won't brake for you, but they'll scream before impact. Given that 2.42 million people were injured last year, the warning alone has value.

Limitations

FARS captures fatal crashes only. We cannot compute model-level injury rates from this dataset. The injury-to-death ratio shift may partially reflect improved trauma care (deaths reclassified as injuries due to faster medical response), not just vehicle engineering. Reporting changes also affect injury counts: states that expanded crash reporting thresholds capture more low-severity incidents. The fleet composition math assumes equal crash frequency across classes, which oversimplifies. Sedans may crash at different rates than SUVs due to driver demographics.

Strongest Counterargument

AEB is already reducing crashes, not just fatalities. IIHS data shows rear-end crash frequency dropping among equipped vehicles, and new-vehicle AEB penetration hit roughly 95% by 2024 model year.^[6] The injury plateau may be a LAG indicator: as more equipped vehicles enter the fleet, injury counts should follow fatalities downward within 5 to 8 years. The story may not be that prevention has failed, but that prevention technology hasn't reached enough of the fleet yet.