Hyundai Wins Every Safety Award. FARS Says It Barely Matters.

I ran every major brand through the same filter: total FARS-linked deaths for model years 2002–2006 vs. 2017–2021. Same methodology, same decade of fatal-crash data, same denominator. Ford dropped 76%. Chevrolet dropped 76%. Chrysler dropped 87%. Hyundai dropped 6%.

That’s not a rounding error. Hyundai’s 2002–2006 model year vehicles accumulated 1,284 fatalities in FARS data. Its 2017–2021 vehicles: 1,212. Seventy-two fewer deaths across five model years and eleven models.^[1] Meanwhile, Hyundai’s own press releases trumpet leading the industry in IIHS TOP SAFETY PICK awards.^[2]

Subaru, another brand synonymous with safety marketing, managed only -21%. Nissan: -28%. The three brands Americans most associate with “safe, affordable imports” are the three worst improvers in the FARS data.

Drill into Hyundai’s lineup and the picture sharpens. The Elantra’s model-year deaths actually increased—from 388 (MY 2002–06) to 498 (MY 2017–21), a 28% rise. The Sentra, Nissan’s budget sibling, went from 512 to 542. Newer versions of these cars are accumulating more fatal-crash involvement than the ones they replaced.

Before you blame the car, acknowledge the denominator: Hyundai’s U.S. sales roughly doubled between those eras, passing 15 million cumulative in 2022.^[3] More cars on roads means more exposure. If deaths held flat while volume doubled, the per-vehicle risk fell around 50%. Reasonable. Except Honda’s U.S. sales stayed roughly flat over the same window and still cut model-year cohort deaths 66%. Toyota sold twice as many models and managed -58%. Volume helps explain Hyundai’s numbers. It does not excuse them.

IIHS and FARS measure different realities.^[4] IIHS tests controlled frontal overlap, side impact, roof strength, and headlight quality in a lab. FARS records every fatal crash on American roads, where speed, impairment, seatbelt noncompliance, and road geometry swamp the marginal advantage of a slightly better crumple zone. A car that aces a 40-mph offset test can still kill its driver at 75 on an unlit rural two-lane.

What You Can Do

IIHS awards tell you how a car performs in a curated set of crash scenarios. FARS tells you how the entire fleet of that car performs in the real world. If you are shopping Hyundai, Subaru, or Nissan, the crash-test stickers are not lying, but they are answering a narrower question than you think. Check the specific model’s FARS death count, not just its award shelf, at nhtsa.gov/FARS. Favor models where newer model years show steep per-unit death declines. Avoid the Elantra and Sentra if the numbers bother you; consider the Civic, Corolla, or Mazda3 instead, all of which tracked the fleet-wide safety curve downward.