Why an old five-star crash rating is not today's five-star

No. A crash rating certifies how a car did on one version of a test in one model year, and both U.S. rating programs have repeatedly made their tests harder. A car that earned five stars or a "Good" in 2010 passed a bar that has since been raised, so the old score tells you less about protection today than the same score on a current model. Compare cars by test version and model year, not by the headline number alone.

That matters most when you buy used. A 12-year-old sedan with "top crash ratings" on a listing was probably a strong performer for its time. Whether it protects you as well as a 2024 car is a separate question the rating cannot answer.

What a crash rating actually certifies

A rating is a measurement under fixed conditions: a particular speed, barrier or pole, type of dummy and set of injury limits. Change any of those and the same car can score differently. Two programs matter in the United States:

• NHTSA's New Car Assessment Program (NCAP), the federal "5-Star Safety Ratings." It tests frontal, side and rollover performance and publishes star ratings. The ratings are consumer information, not a legal standard a car must meet.
• The Insurance Institute for Highway Safety (IIHS), funded by insurers, which rates vehicles Good, Acceptable, Marginal or Poor in separate tests and gives annual Top Safety Pick awards.

Neither tells you whether a specific car on the lot is in good condition. A rating assumes the vehicle left the factory as tested. It does not account for prior crash repairs, replaced airbags or missing parts, which is why a vehicle history report and an inspection belong in any used-car purchase.

How NHTSA's ratings have changed

NHTSA's own history shows how often the program moved. In a September 2025 Federal Register notice, the agency said it established NCAP in 1978 and began frontal impact testing in 1979. It added side-impact barrier testing in 1996, rollover assessments in 2000 and 2003, and side-impact pole testing and driver-assistance assessments beginning with model year 2011 (Federal Register, September 22, 2025).

The model year 2011 overhaul was the biggest shift. According to NHTSA's December 2024 notice, the last major upgrade was published July 11, 2008 and took effect with 2011 vehicles. It introduced new crash test dummies, an oblique pole test and an overall rating that combines frontal, side and rollover results. NHTSA said at the time that the new ratings used different test conditions and injury criteria and could not be compared with ratings of earlier model years.

That is a point worth sitting with. A car that scored five stars for model year 2009 and a car with five stars for 2013 were scored against different yardsticks, and NHTSA does not present them as equivalent. Lower star counts after the change did not mean the cars had become less safe; they meant the test had become more demanding.

Driver-assist checkmarks, and the 2026-27 delay

NHTSA also tracks crash-avoidance technology, but with checkmarks rather than stars. According to the September 2025 notice, it assigns stars for the original program elements (frontal, side and rollover) and uses checkmarks to recommend technologies that meet its performance requirements. It added electronic stability control, forward collision warning and lane departure warning in 2011, rear visibility systems in 2013, and automatic emergency braking features (crash imminent braking and dynamic braking support) for model year 2018.

In December 2024 NHTSA announced four new technologies (blind spot warning, blind spot intervention, lane keeping assist and pedestrian automatic emergency braking), tougher AEB criteria and a new pedestrian crashworthiness program, all to begin with model year 2026. In September 2025 it postponed those changes by one model year, to model year 2027. As of November 2026, then, the expanded elements formally apply to 2027 vehicles. NHTSA said it would still give credit to 2026 models whose manufacturers attested to passing the new tests, so a 2026 car's checkmarks may reflect either set of criteria.

A weight-class rule many shoppers miss

NHTSA's frontal crash star rating simulates a crash between two vehicles of similar weight. For that reason NHTSA says to compare frontal ratings, and the overall score that depends on them, only with vehicles in the same class and within about 250 pounds of each other (see NHTSA's ratings page). Side and rollover results do not carry that limit. A five-star compact is not "as safe" as a five-star full-size pickup in a frontal crash; the stars describe each against its own weight peers.

How IIHS tests have changed

IIHS's test list is a record of tests becoming harder after cars learned to pass the old ones.

Sources: IIHS releases of August 2012, October 2021 and December 2022.

Two lessons come out of that table.

Rankings reset when the test changes. The December 2022 release found that all 15 small SUVs earned Good in the original moderate overlap test but only the Ford Escape and Volvo XC40 earned Good in the updated one, because most did not adequately protect a rear passenger's head and neck. IIHS also adjusted the updated test's rating criteria in 2024 to account for variations in rear shoulder belt position, which changed some overall ratings (the release says the Toyota RAV4's moved from Acceptable to Marginal). Even a published rating can be re-scored.

A Good in an old test sits beside a worse score in a new one. IIHS publishes both when a vehicle has been evaluated twice. On its vehicle pages, you may see an "original" and an "updated" rating for the same model. Do not stop at the better-looking one.

"Good" now means more than "Good" then

IIHS's award criteria tighten too. For 2026, according to its March 24, 2026 release, a Good rating in the updated moderate overlap test is required for either Top Safety Pick award (acceptable was enough for the base award in 2025), and Top Safety Pick+ requires Good in the pedestrian front crash prevention test and Acceptable or Good in the new vehicle-to-vehicle test. Qualifying front crash prevention must be standard on every trim, and some winners count only if built after a stated date (IIHS).

The practical point: a "Top Safety Pick" badge from five or ten years ago was measured against a smaller checklist. A used car that carried the badge in 2016 was never tested for the back-seat dummy, the heavier side barrier or the pedestrian tests that a 2026 winner passes.

Ranking within a year versus measuring against a standard

Ratings behave differently, and shoppers often assume they are one or the other:

• Star ratings tie injury risk to a fixed scale for that program version. NHTSA converts measurements from the dummies into a probability of injury and then into stars. But because the 2011 program changed the dummies and the criteria, the stars from different eras are not on one continuous scale, and frontal stars are only comparable within a weight class.
• IIHS ratings use thresholds on structure and dummy readings. A vehicle needs to meet limits for Good, regardless of how other cars did. That is closer to a fixed standard, but IIHS sets the standard, so it moves when IIHS changes the test.
• Awards are partly relative. Top Safety Pick criteria are set each year, and the list of winners reflects what manufacturers have built.

In every case, the rating is meaningful only inside its own test version and year.

Do ratings predict real injuries?

Yes, though not perfectly. A study by IIHS researchers Eric Teoh and Samuel Monfort in Traffic Injury Prevention (June 2023) compared driver deaths per police-reported frontal crash by small overlap rating. After adjusting for vehicle type, weight and driver demographics, drivers of vehicles rated Good were 12% less likely to die in frontal impacts than drivers of Poor-rated vehicles; Acceptable-rated vehicles showed an 11% reduction, and the 5% difference for Marginal was not statistically significant (IIHS bibliography). The test was created in 2012 because virtually all vehicles were earning Good in the long-running moderate overlap test while serious frontal crashes continued.

That finding supports the idea that tougher tests find real differences. It also shows limits: the reduction is modest, federal crash databases cannot isolate small-overlap crashes, and the study covers drivers only. Insurers use a different real-world check, claim data. Our guide to IIHS and HLDI insurance loss data: what it shows about cars explains the Highway Loss Data Institute's injury and collision claim frequencies by make and model, which can contradict what a crash test suggests.

How to find the test year and version

Use this process before trusting a rating on a used-car listing:

1. Get the exact model year and body style. Ratings are issued per model year and often per body style (sedan versus hatchback, two-wheel versus four-wheel drive).
2. Look up NHTSA's page for that year. Check whether it is a 2011-or-later rating and read the frontal, side and rollover pieces separately. Note the vehicle's weight class if you plan to compare frontal results.
3. Look up IIHS's page for the same model year. Read each test: small overlap front, moderate overlap (original or updated), side (original or updated) and, if listed, headlights and front crash prevention.
4. Check the build date. IIHS notes when a result applies only to cars built after a certain month because the manufacturer changed something mid-year, such as the Escape's rear belts (built after May 2022).
5. Note what is missing. A blank next to "updated moderate overlap" or "updated side" on a car that is a few years old means it has not been put through the current test.
6. Look at the equipment, not just the badge. Whether automatic emergency braking and other driver-assist systems were standard, optional or absent matters as much as crash structure on modern cars. IIHS has found that automatic emergency braking cuts rear-end crashes roughly in half, so a car lacking it is behind regardless of its crash stars.
7. Check recalls and repair history. See our article on vehicle recalls and insurance: what an open recall changes and the used-car steps in insuring a used car.

A worked comparison

Say you are choosing between two midsize SUVs. Car A is a 2012 model whose listing says "5-star NHTSA, IIHS Good frontal." Car B is a 2024 model with similar claims. Here is how to compare them, with the specific findings clearly hypothetical:

Both show "five stars," but only one has been through the tests that the industry has had to meet since. If you want Car A for price reasons, the missing rows are the questions to weigh, not a reason for panic: many older cars are reasonably protective, and the gap is largest for rear passengers and for side crashes with modern SUVs.

What this means for insurance

Star ratings are not a rating factor on most policies. Insurers price on claim experience for the model (a vehicle symbol) and on the other rating factors that describe you, your location and your coverage. Cars with strong structural protection and crash-avoidance technology tend to have fewer or less costly injury claims, but the same technology can raise repair costs; see driver-assistance sensors and the cost of repairs. For related background on how a car's design changes its premium, see how a redesign changes a car's insurance rating group and your premium and our guide to vehicle safety ratings and insurance costs. Get a quote on the exact car before you buy, since a newer, better-rated vehicle can cost more or less to insure than an older one.

Common questions

Is a 2008 five-star car safer than a 2012 four-star car?

You cannot tell from the stars. NHTSA changed the program for model year 2011 and said the new ratings are not comparable with earlier ones. Compare the specific tests, equipment and, where possible, real-world data for each model.

Do IIHS awards from years ago still count?

An award reflects the criteria for that year. Check the current IIHS page for the model year rather than relying on a badge on an old listing, because the requirements for 2026 are stricter than for earlier years.

Why would a car lose a rating I saw last year?

Either the test changed (as with the updated side and moderate overlap tests) or IIHS adjusted its scoring criteria, as it did in 2024. The car did not change; the yardstick did.

Does a higher star rating lower my premium?

Not directly. Insurers use claim data and rating plans, not star counts. A safer car can still be expensive to repair. Compare quotes for the exact vehicle.

Key takeaways

• A crash rating is tied to a specific test version and model year. NHTSA's model year 2011 overhaul and IIHS's 2012, 2021 and 2022 test updates all made the bar harder.
• NHTSA frontal star ratings should only be compared within the same class and within about 250 pounds; stars from before and after 2011 are not comparable.
• When IIHS shows "original" and "updated" results for the same car, the updated one is the more demanding test; a car that is Good in one and weaker in the other may still be a reasonable choice, but know the difference.
• Ratings track real-world risk: in a 2023 study, Good small overlap ratings were associated with 12% lower frontal-crash driver death risk than Poor ratings.
• Before buying used, check the exact year and body style, the build date, driver-assist equipment, recalls and repair history, and get an insurance quote.