What IIHS and NHTSA Safety Ratings Actually Tell You About an EV

Published August 10, 2026

Two different organizations put “safety ratings” on new vehicles in the U.S., and they’re not testing the same things the same way. Mixing them up, or assuming a 5-star EV and a “Good”-rated EV are directly comparable across models, is an easy way to draw the wrong conclusion. Here’s how each one actually works, and what’s worth knowing if the vehicle in question is electric.

IIHS: an insurance-industry-funded nonprofit with its own test protocols

The Insurance Institute for Highway Safety is a nonprofit funded by U.S. and Canadian auto insurers, not a government agency. It runs its own crash tests, separate from and generally more demanding than the federal government’s, and rates vehicles on two dimensions: crashworthiness (how well the structure and restraints protect occupants in a crash) and crash avoidance and mitigation (technology that prevents a crash or reduces its severity).

Crashworthiness tests include a moderate-overlap frontal test (40% of the front end hitting a similarly sized vehicle at 40 mph), a small-overlap frontal test on both the driver and passenger side (mimicking a hit against a tree, pole, or the corner of an oncoming car), a side-impact test, roof strength, and head restraint evaluation. IIHS rates each test Good, Acceptable, Marginal, or Poor. In February 2024, IIHS updated the scoring criteria for its moderate-overlap test with a new “chest index” metric that accounts for rear-seat shoulder-belt position and chest compression, since research had found rear-seat occupants were, counterintuitively, now facing a relatively higher fatal-injury risk than front-seat occupants in newer vehicles (a byproduct of front seats getting safer faster than rear seats, not rear seats getting worse).

To earn a 2026 Top Safety Pick award, a vehicle needs Good ratings in the updated moderate-overlap test, side-crash test, and both small-overlap tests, plus Acceptable-or-better ratings in pedestrian front-crash prevention and headlight performance, with front crash-prevention tech standard. Top Safety Pick+ is new and stricter for 2026: it now requires a Good rating in the pedestrian test and an Acceptable-or-better rating in an updated vehicle-to-vehicle front crash-prevention test that includes higher-speed scenarios against motorcycles and semi-trailers.

NHTSA: the federal 5-star program

The National Highway Traffic Safety Administration is a U.S. Department of Transportation agency, and its New Car Assessment Program is the source of the 5-star ratings you’ll see on window stickers and NHTSA’s public database. It tests three things: a frontal crash (a 35 mph run into a rigid barrier, using both a 50th-percentile male and 50th-percentile female crash-test dummy), a side crash (a moving barrier hitting the vehicle at 38.5 mph), and rollover resistance (a calculated static stability factor rather than an actual rollover test).

Each star roughly corresponds to a 10% change in the probability of a serious injury: a 5-star rating means less than a 10% chance of major injury for the driver or passenger, while a 1-star rating means at least a 40% chance. The detail that trips people up most: NHTSA ratings are only meant to be compared within the same weight class, vehicles of similar size and curb weight. A 5-star subcompact and a 5-star full-size SUV aren’t equally protective in an actual crash between the two; the rating tells you how a vehicle protects its own occupants relative to peers its size, not how it would fare against something much heavier or lighter.

NHTSA had planned to add new crash-avoidance categories, pedestrian automatic emergency braking, lane-keeping assist, and blind-spot intervention, to the 5-star program starting with 2026 models, but that overhaul was delayed after an industry request, and is now expected to begin with 2027 models instead. If you’re comparing 5-star ratings on 2026 model-year vehicles today, you’re still looking at the older, narrower version of the program.

What’s actually different about EVs

The short answer: EVs are tested the same way gas vehicles are, but their construction changes some of what shows up in results. A big, flat battery pack mounted low in the floor lowers the center of gravity, which generally helps rollover resistance. It also adds significant curb weight (often several hundred to well over a thousand pounds versus a comparable gas model), which changes crash dynamics for both the EV’s own occupants and the occupants of whatever it hits.

IIHS data covering 2011-2019 model years found injury claim rates for EV occupants were more than 40% lower than for occupants of comparable gas-powered versions of the same model, evidence that the added mass and low center of gravity have generally worked in EV occupants’ favor. But that same mass cuts the other way for the other vehicle in a collision: in tests measuring rear-seat occupant risk, IIHS has flagged specific concerns in individual EVs (elevated chest-injury risk from rear seat-belt forces in some models), a reminder that “Good” or “5-star” ratings summarize a lot of individual test measurements and it’s worth reading past the headline grade if you want the full picture on a specific vehicle.

Weight has also become a literal engineering problem for the testing organizations themselves: IIHS has noted that some of the heaviest EVs and trucks on the market, in the 9,000+ pound range, push against the limits of its existing crash-test equipment, originally designed around lighter gas-vehicle fleets. On the fire question specifically, which tends to dominate public perception of EV crash safety, NHTSA’s own review has concluded that EVs do not present a greater risk of post-crash fire than gasoline vehicles, and the agency runs a dedicated Battery Safety Initiative to keep studying battery-specific crash and fire risks as the vehicle fleet evolves.

Reading the two ratings together

Neither rating alone tells the whole story. IIHS tests are generally more demanding and more frequently updated, and its Top Safety Pick criteria increasingly reward advanced crash-avoidance tech, useful if you want to know how a vehicle stacks up against current best practice. NHTSA’s 5-star system is the free, government-run baseline most car shoppers see first, but its weight-class caveat and its currently-delayed technology overhaul mean it’s a narrower snapshot than it looks. For an EV specifically, a strong rating from either body is a reasonable signal that the battery pack’s weight and placement haven’t compromised occupant protection, but it’s not a substitute for checking a specific model’s individual test results if towing, hauling, or a mixed fleet of vehicle sizes on your household’s roads matters to you.

Sources

Crash-test protocols and award criteria are updated periodically by both organizations; treat specific pass/fail thresholds as accurate as of the 2026 model year and worth rechecking for future model years.