Understanding Your Vehicle’s Safety Ratings

Two organizations crash-test most new cars sold in the United States, and they don't measure the same things or report results the same way. A car can earn five stars from one and a weak score from the other. Knowing what each one tests makes it much easier to compare models and to spot the gaps a headline rating can hide.

NHTSA: the government's five-star ratings

The National Highway Traffic Safety Administration runs the New Car Assessment Program, which publishes the familiar one-to-five-star ratings. It is a federal agency, and its scores appear on the window sticker of new cars.

Frontal crash

The car is driven head-on into a rigid barrier at 35 mph with crash-test dummies in both front seats. Instruments measure the forces on the head, neck, chest and legs, and those readings are turned into a star rating for the driver and the passenger.

Side crash

A barrier is pushed into the side of the stationary car to imitate being hit at an intersection, and a second test slides the car sideways into a pole, the kind of crash that happens when a car leaves the road.

Rollover resistance

This one is mostly a calculation based on how top-heavy the vehicle is, combined with a driving test. It produces an estimated chance of rolling over in a single-vehicle crash, which is why tall SUVs and pickups often score lower here than sedans.

The overall star rating combines the three. Stars are most meaningful when you compare vehicles of similar size and weight, because the frontal test is crashed into a fixed barrier and doesn't capture what happens when a small car meets a large one.

IIHS: the insurance industry's tougher tests

The Insurance Institute for Highway Safety is an independent, nonprofit research group funded by auto insurers. It rates results as Good, Acceptable, Marginal or Poor, and it regularly makes its tests harder as carmakers catch up.

Small overlap front

Introduced in 2012, this test puts only about a quarter of the car's front width into a barrier, imitating a collision with a pole, a tree or the corner of another vehicle. It bypasses much of the main crash structure, and many cars that did well in traditional tests performed badly when it first appeared.

Updated moderate overlap front

The original version measured protection for front-seat occupants. The updated test, which arrived in 2022, adds a dummy sized like a small woman or a 12-year-old in the back seat. Rear-seat protection has lagged behind front-seat safety in many models, and this is where it shows.

Updated side test

IIHS replaced its original side test with a heavier barrier, about 4,200 pounds, moving faster, at roughly 37 mph. The change reflects the fact that the vehicles most likely to hit you from the side today are SUVs and pickups.

Headlights and crash avoidance

IIHS also measures how far a car's headlights actually illuminate the road without dazzling oncoming drivers, and how well automatic emergency braking avoids or slows down for vehicles and pedestrians. Headlight performance can vary by trim level on the same model, which surprises a lot of buyers.

Top Safety Pick awards

Each year IIHS names Top Safety Pick and Top Safety Pick+ winners. The criteria change annually, so an award from five years ago doesn't mean the car would win under today's tougher tests. When you compare models, look at the year attached to the award.

What the tests don't cover

Crash tests measure how a car protects the people inside during a handful of standard collisions. They say less about how easy a car is to live with safely. IIHS separately rates how easy the LATCH anchors for child seats are to use, and parents who have wrestled a car seat into a cramped back seat know how much that matters. It also rates seat belt reminders, since an unbelted occupant gets little benefit from a five-star structure.

Neither organization rates visibility in any detail, and thick roof pillars that help in a rollover can create real blind spots at junctions. A test drive in the kind of traffic you normally face is the only way to judge that.

Real-world death rates

Every few years IIHS publishes driver death rates by model, based on how many drivers died in crashes per million registered vehicles over several years. These figures reflect real roads and real drivers, so they are influenced by who tends to buy each model and how it's driven, but they're a useful check on lab results. The gap between the safest and least safe models is large, and the lowest rates are dominated by larger, newer vehicles with strong test scores.

How to use the ratings when shopping

Start with the overall scores, then open the detailed results for the cars on your shortlist. Look for weak spots: a five-star car might still rate Marginal in the IIHS small overlap test, or have headlights rated Poor on the trim you're considering. Check whether the rating applies to the exact model year you're looking at, because a redesign can change results significantly. For used cars, both organizations keep archives going back many years.

Size still matters in crashes between vehicles. In a collision between a small car and a large SUV, the occupants of the heavier vehicle usually fare better, regardless of how each scored on its own. A well-rated small car is very safe for its size, but comparisons across very different size classes should be made with that in mind.

Both sets of results are free to search: NHTSA at nhtsa.gov/ratings and IIHS at iihs.org/ratings. Each includes crash-test video, which often tells you more in thirty seconds than the summary score does.