31 Cars Almost Never Kill Their Occupants. You Can Buy 28 of Them Today.

Before you sign that lease, you might want to see this. Buried in ten years of FARS fatality data sits a cluster of 31 vehicles with occupant death rates below 0.15 per 100 million vehicle miles traveled.^[1] That is not a rounding error. That is a solved engineering problem sitting in dealership lots across the country while buyers wander in blind and drive off in something 200 times more dangerous.

The club's most absurd member is the Ram 3500, which recorded zero occupant deaths across a 437,500-vehicle fleet over ten years of exposure without a single fatal crash involving anyone inside. The physics explanation is blunt: the truck weighs 7,000 pounds, and physics does not negotiate. But mass alone does not explain the rest of the list, because sitting three spots down is a discontinued Toyota hatchback that weighs half as much.

The Toyota Matrix posted a rate of 0.02 with just six deaths across its entire fleet.^[1] Discontinued in 2014. A compact hatchback that nobody would call a safety icon outperformed every luxury sedan in the FARS database. That fact should bother the entire automotive marketing industry, but it won't, because the Matrix never had a Super Bowl ad.

Who Made the Cut

Twenty-one of the 31 vehicles are SUVs. That is 68 percent of the near-zero cluster dominated by a single body style, and it tracks with what IIHS has documented about mass, ride height, and crash geometry for two decades.^[2] Four are pickups, four are sedans, and two are vans, but the SUV dominance is not accidental; these crossovers combine mass with modern crash structures, standard electronic stability control, and curtain airbags that deploy in configurations some sedans still treat as optional packaging.

The Tesla Model Y sits at 0.03 deaths per 100M VMT with 57 total fatalities across a fleet of 1.75 million vehicles, making it the largest fleet in the club by a wide margin.^[1] The Kia Telluride and Hyundai Palisade both land at 0.04 and 0.06 respectively, having reached near-zero status faster than any midsize SUVs in FARS history despite arriving only in 2020. The Kia Seltos, a subcompact crossover that costs under $30,000, matches the Telluride at 0.04.

Meanwhile, the Hyundai Veloster posts a rate of 8.54, the Ford Mustang lands at 6.02, and the Chevrolet Cobalt, a car GM should have killed before it killed anyone else, sits at 5.10.^[1] The Veloster's rate is 284 times the Model Y's, a gap so grotesque it deserves to be repeated: not double, not triple, not an order of magnitude, but two hundred and eighty-four times higher. These vehicles share the same roads, the same speed limits, the same decade of data collection, and yet one might as well be a bicycle on a freeway while the other is a bunker on wheels.

The Uncomfortable Math

Vehicle class aggregates tell the structural story. Sports cars as a class post 26.88 deaths per 100M VMT. Sedans: 13.92. Pickups: 8.86. Vans: 6.54. SUVs: 5.47.^[1] The near-zero club averages roughly 0.08. That makes these 31 vehicles 69 times safer than the sports car class aggregate and 174 times safer than the deadliest individual models in the database.

This is where honest methodology demands a caveat, so here it is at full strength: newer vehicles with fewer accumulated vehicle miles traveled will mechanically produce lower observed death rates because they have had less exposure time, because survivors tend to drive newer and better-maintained vehicles, and because the people buying Tellurides and Palisades may simply be more cautious drivers than the people buying Mustangs and Velosters.^[3] Selection bias is real and probably inflates the gap. But the Toyota Matrix undermines the newness argument entirely, having achieved near-zero despite being a discontinued econobox with over a decade of accumulated exposure, and the Ram 3500's zero-death record spans models from 2014 through 2023. Exposure time alone cannot explain a fleet of 437,500 trucks that never killed an occupant.

What This Means If You Are Shopping Today

You can buy your way into the near-zero club for under $30,000, because the Kia Seltos starts at $24,590 and the Hyundai Tucson, the Mazda CX-5, the Subaru Forester, and the Honda CR-V all appear in or near the threshold. None of these are exotic or require a luxury badge. They are boring, mid-trim crossovers sitting on every dealer lot in America, and their occupants almost never die in crashes.

If you are currently driving a Chevrolet Cobalt, a Nissan Maxima, a Hyundai Veloster, or a Ford Mustang, your vehicle's occupant death rate is between 34 and 284 times higher than the vehicles listed above, and that is not a scare tactic but a FARS spreadsheet.

Limitations

FARS captures only fatal crashes, which represent a fraction of the roughly 6.7 million annual U.S. crashes. A vehicle with a low fatality rate may still have high injury rates that this data cannot measure. Estimated death rates use VMT approximations rather than actual odometer readings, introducing uncertainty of approximately ±15% for lower-volume models. Driver demographics, usage patterns, and geographic concentration all confound direct vehicle-to-vehicle comparisons. The near-zero cluster is a signal, not a controlled experiment.