← The Crash Report
The Gap

America’s Favorite Vehicle Class Has a 14x Death Rate Spread. Nobody Tells You Which End You’re On.

☕ 5 min read
A row of compact crossover SUVs in a dealership lot, some casting long shadows while others are brightly lit, representing the hidden safety gap between seemingly identical vehicles

Walk into any dealership in America and you will find a compact crossover SUV. Probably six of them. They all look vaguely similar: raised ride height, chunky plastic wheel arches, a liftgate you can open with your foot if you paid for that package. They cost roughly the same. They weigh roughly the same. And if you picked randomly, you might be 14.4 times more likely to die in one than another.

14.4×
Death rate spread between the safest and deadliest compact SUVs in FARS data (2014–2023). Same vehicle class. Nearly identical impairment rates.

That number comes from ranking 13 of America’s top-selling compact crossovers by FARS fatality rate per 100 million vehicle miles traveled. At the bottom: the Mazda CX-5, at 0.12 deaths per 100M VMT. At the top: the Jeep Cherokee, at 1.73.[1] Both classified as compact SUVs. Both priced in the low-to-mid $30,000s. Both driven by people with functionally identical impairment profiles.

Same Drivers, Wildly Different Outcomes

Before anyone blames the usual suspect: no, this is not a drunk-driving story. Impairment rates across all 13 models fall within a 3-percentage-point band, from 17.6% (CR-V and Tucson) to 20.5% (Forester). That’s statistical noise. Whatever separates a CX-5 fatality from a Cherokee fatality, it is not the blood alcohol content of the person behind the wheel.

Here is how they stack up, ranked safest to deadliest:

Mazda CX-5: 0.12 • Toyota RAV4: 0.19 • Subaru Forester: 0.26 • Kia Sportage: 0.28 • GMC Terrain: 0.29 • Hyundai Tucson: 0.34 • Nissan Rogue: 0.35 • Chevy Equinox: 0.36 • Mitsubishi Outlander: 0.37 • Hyundai Santa Fe: 0.39 • Honda CR-V: 0.53 • Ford Escape: 0.95 • Jeep Cherokee: 1.73

Read that list again. Notice the cliff between the Outlander at 0.37 and the CR-V at 0.53? A 43% jump. Then the Escape nearly doubles the CR-V. Then the Cherokee nearly doubles the Escape.

The Sacred Cow Problem

The Honda CR-V is the default recommendation in every car-buying guide published since 2014. Consumer Reports loves it. Your parents love it. Your parents’ financial advisor who drives one also loves it.

At 0.53 deaths per 100M VMT, the CR-V ranks 11th out of 13 compact SUVs. It is 4.4 times deadlier than the CX-5, and 2.8 times deadlier than the RAV4 it’s constantly compared against. One external analysis found the 2019 CR-V was 18% more likely to result in a fatal accident compared to other SUVs.[2] Not terrible in the grand scheme of vehicles. Middling for its class. But nowhere near the safety halo its reputation implies.

Is 0.53 dangerous? In absolute terms, no. Compared to a sedan averaging 1.67, it’s excellent. But that’s a category-level compliment. Within the compact SUV class, you can do 4x better for the same money.

What Is Wrong with the Cherokee?

A death rate of 1.73 in a compact SUV is abnormal. For context, the average sedan death rate across 67 models is 1.67. The Jeep Cherokee, despite the mass and ride-height advantages that make SUVs structurally safer than sedans, manages to be deadlier than the typical sedan. That takes effort.

NHTSA opened an investigation (PE23-012) into 2014–2022 Cherokee KL models over electronic parking brake failures and engine stalling.[3] The KL platform’s ZF 9-speed automatic transmission was a documented reliability nightmare, with complaints about rough shifting, hesitation during acceleration, and loss of power. Stellantis discontinued the Cherokee in 2023.

At 2,276 deaths over the FARS observation period and a fleet of 1,050,000 vehicles, the Cherokee’s body count is comparable to the Ford Escape’s 2,284, but the Escape has nearly double the fleet size (1.925 million). More vehicles, roughly the same deaths. That means the Escape’s per-mile rate (0.95) is bad but at least proportional. The Cherokee’s rate (1.73) is disproportionate.

Who Actually Wins

Japanese and Korean engineering dominates the top half. The CX-5 at 0.12, the RAV4 at 0.19, and the Forester at 0.26 form an upper tier so far ahead of the field that you could pick any of the three blindfolded and still be driving one of the safest vehicles on American roads, period. Not just safest for its class. Safest overall. An RAV4 at 0.19 is safer per mile than 95% of all vehicles in the FARS database, regardless of size or price.[1]

The GM entries (Terrain at 0.29, Equinox at 0.36) land solidly mid-pack, which for GM is actually an achievement worth noting.

What This Does Not Prove

FARS captures only fatal crashes. The 2,072 CR-V deaths are a fraction of the millions of CR-V crashes that did not kill anyone. A vehicle with a low fatality rate might still have high injury rates, and FARS cannot tell us that. Our VMT-based rate estimates carry roughly ±15% uncertainty for models with smaller fleets. The Outlander’s 0.37 rate is based on a fleet of 393,750, making its confidence interval wider than the RAV4’s, which draws from 3.76 million vehicles.[4]

Fleet age composition also matters. The Cherokee KL arrived in 2014 and ended production in 2023, meaning the average Cherokee on the road during the FARS observation window is older than the average CX-5 (continuous updates since 2012, major redesign in 2017). Older vehicles lack newer active safety tech.

But the RAV4 is a similar vintage, and it scores 0.19.

What This Means for You

If you are shopping for a compact SUV, you are already making a statistically sound decision. As a class, compact crossovers are safer than sedans, pickups, and sports cars. But within that class, the model you choose can shift your risk by an order of magnitude. A CX-5 buyer and a Cherokee buyer are both buying “a compact SUV.” They are not buying the same level of protection.

IIHS star ratings hint at this, but they grade crashworthiness in controlled lab conditions.[5] FARS data shows what happens on actual roads, over billions of actual miles, driven by actual people who were statistically identical in their drinking and drug habits.

The brochure won’t tell you. The salesperson won’t tell you. Now you know.

Sources & References

  1. NHTSA, Fatality Analysis Reporting System (FARS), 2014–2023. Per-model death rates calculated from FARS fatal crash records, estimated fleet sizes from industry sales data, and VMT estimates from NHTS annual mileage by vehicle type. nhtsa.gov
  2. RealSafeCars.com, 2019 Honda CR-V Safety Analysis. Reports CR-V is 18% more likely to result in a fatal accident compared to other SUVs, with a safety score of 75/100. realsafecars.com
  3. NHTSA, Preliminary Evaluation PE23-012, investigation into 2014–2022 Jeep Cherokee KL electronic parking brake and engine stalling complaints. nhtsa.gov
  4. NHTS (National Household Travel Survey), annual vehicle miles traveled estimates by vehicle type. Fleet size estimates derived from cumulative U.S. sales data and assumed 15-year fleet life. nhts.ornl.gov
  5. IIHS, Small SUV Ratings, crashworthiness and crash avoidance evaluations for the compact SUV class. iihs.org

Source: NHTSA FARS 2014–2023. Death rates are per 100 million vehicle miles traveled, estimated from fleet size and NHTS average annual mileage. Impairment percentages reflect BAC > 0 or drug-positive toxicology in FARS-reported fatal crashes involving that vehicle’s driver. See methodology for caveats.