← The Crash Report
By The Numbers

A Single Number Exposes America’s Most Lethal Used Cars

I built a metric nobody asked for. Two inputs: how often a vehicle appears in a fatal crash per fleet unit, and how often its own occupant dies when it does. Multiply them. What falls out is a single number that captures both pathways to killing you at once.[1]

652 vs. 6
Composite Danger Score: Chevy S-10 vs. Tesla Model 3. A 100-to-1 spread.

A Chevrolet S-10 pickup scores 652 on this composite Danger Score. A Tesla Model 3 scores 6. That is a 100-to-1 spread across both dimensions of real-world risk, sitting in a gap that no existing safety metric captures. NHTSA publishes death rates per VMT. IIHS runs lab crash tests. Insurance actuaries compute claims frequency. Nobody multiplies the two independent axes of danger into a single number.[2]

Crash frequency: fatal involvements per 100,000 fleet vehicles, 2014 through 2023. Nissan's Maxima leads at 888, the S-10 at 864, then a cliff down to Tesla's Model 3 at 13.5. Crash lethality: when this vehicle IS in a fatal crash, how often does its occupant die? A Cavalier loses 85.7% of the time. A Cobalt, 80.8%. A Ford F-250, only 36.4%.[1]

Multiply, divide by 100, rank. Five worst composites among vehicles with 100K+ fleet and 200+ fatal involvements:

1. Chevrolet S-10: 652. Frequency 864, lethality 75.5%. A compact pickup from when "compact" meant "no crumple zone and the frame rails point at your spleen."
2. Nissan Maxima: 588. Frequency 888, lethality 66.2%. Nissan marketed it as a four-door sports car. FARS confirmed it.
3. Chevrolet Cobalt: 587. Frequency 727, lethality 80.8%. GM's ignition switch scandal killed 124 people through a known defect, but the Cobalt's problems run deeper than one faulty part.[3]
4. Chevrolet Impala: 575. Frequency 852, lethality 67.5%. Rental lots saturated America with aging Impalas that kept crashing long after the rental companies moved on.
5. Toyota Solara: 489. Frequency 714, lethality 68.5%. Inherited Toyota's reliability but not its crash protection. You could drive it forever; whether you survived was another question.

Five safest: Tesla Model 3 (5.9), Mazda CX-5 (15.4), Honda HR-V (16.2), Ram 1500 (17.0), Toyota RAV4 (24.3). Model 3 gets there through frequency alone, barely showing up in FARS at 13.5 per 100K. Ram 1500 takes the opposite route: crashes at a moderate 49.9 per 100K but kills its own occupant only 34.1% of the time. Two survival strategies, comparable composite safety, opposite mechanisms.

Why Nobody Publishes This

Fair reasons exist. A Cobalt doesn't score 587 purely because of engineering. Cobalt drivers skewed young, low-income, piloting aging examples with deferred maintenance on roads without median barriers. What the composite actually measures is the full ecosystem of risk around a vehicle, not its structural performance in a lab.[2]

That is also why it's useful. You don't buy a crash test rating. You buy the real-world package: the vehicle, the roads, the maintenance history of a ten-year-old example, the statistical company you'll keep in its fleet. IIHS Top Safety Pick tells you what happens in a controlled collision. This tells you what happens when you actually drive the thing home.

Strongest Case Against

Vehicle age. Every worst-scorer is old: S-10 discontinued 2004, Cobalt ran 2005-2010, Maxima's worst generations are early 2000s. Older fleets pile up both higher frequency (degraded machines, fewer safety features) and higher lethality (pre-ESC, fewer airbags). Maybe the composite just measures how old the surviving fleet is.

I tested this. Among 2018+ models with sufficient FARS data, the best-to-worst spread still exceeds 20:1.[1] Age amplifies the signal. It doesn't create it.

If You’re Shopping

Avoid body-on-frame compact pickups from the pre-ESC era, subcompact GM sedans from the cost-cutting decade (Cobalt, Cavalier, Ion), and high-volume fleet sedans with known frequency problems (Impala, Maxima). Check both axes independently: BMW's 3 Series crashes at 529 per 100K with 59.4% lethality for a composite of 314, while Kia's Forte manages 68.9 per 100K and 66.8% lethality for a composite of 46. Three times the price, seven times the danger.[1][2]

Check your VIN at nhtsa.gov/recalls before driving anything on the worst list. Then, honestly, consider not driving it at all.

Sources & References

  1. NHTSA, Fatality Analysis Reporting System (FARS), 2014–2023. Crash frequency, lethality, and fleet data derived from FARS bulk CSV cross-tabulated with US vehicle sales estimates and NHTS annual miles. nhtsa.gov
  2. IIHS, Fatality Statistics: Yearly Snapshot. Comparative context for per-model fatality rates and crashworthiness ratings. iihs.org
  3. General Motors ignition switch recalls (2014). 124 deaths attributed to faulty ignition switch in Cobalt and related models. Wikipedia

Source: NHTSA FARS 2014–2023. Composite Danger Score = (fatal crash involvements per 100K fleet) × (occupant death rate per involvement) / 100. Fleet estimates use sales-derived proxies with ±15% uncertainty for low-volume models. See methodology for caveats.