The moderate overlap front test has existed for decades. A car hits a barrier with part of its front end, and the instruments read what happened to the driver.
In 2022 IIHS added a second dummy, the size of a small woman or a twelve-year-old, in the seat behind the driver. Nothing else changed.
The results collapsed.
| Vehicles rated under both versions | Count |
|---|---|
| Rated Good on the original test | 227 |
| Still Good with a rear dummy | 58 |
| Dropped to Poor or Marginal | 147 |
Two thirds of the cars that had a top score lost it, and they lost it without a single change to the vehicle. The car did not get worse. The question got harder.
What the rear dummy actually measures
The front seat has had thirty years of engineering attention. Load-limiting belts that pay out under high force, pretensioners that pull the slack before the crash develops, airbags shaped around a specific occupant position.
Most of that never migrated to the back. A rear belt is often a simple three-point without a load limiter, and in the updated test that shows up as chest forces high enough to indicate rib injury, and as the dummy sliding underneath the lap belt.
The finding that put this test on the map was that the back seat had become the more dangerous place to sit in a modern car — not because it got worse, but because the front got so much better.
The brands that struggle most
Among brands with at least fifteen vehicles rated on the updated test, the failure rate is high almost everywhere, including brands with strong safety reputations.
| Brand | Rated | Poor or Marginal |
|---|---|---|
| Jeep | 32 | 26 |
| Lexus | 15 | 12 |
| Toyota | 53 | 37 |
| Chevrolet | 30 | 22 |
| Subaru | 34 | 21 |
| Ford | 40 | 23 |
Lexus and Toyota appearing here matters more than Jeep does. These are brands that score well on the star ratings, and their presence shows this is not a cheap-versus-expensive problem. It is a front-versus-rear problem, and almost nobody had solved it.
It is being fixed fast
Unlike most safety findings, this one is moving in the right direction quickly.
| Model year | Rated | Poor or Marginal | Share |
|---|---|---|---|
| 2022 | 39 | 30 | 77% |
| 2023 | 77 | 54 | 70% |
| 2024 | 112 | 55 | 49% |
| 2025 | 150 | 60 | 40% |
| 2026 | 147 | 46 | 31% |
From 77% failing to 31% in four model years. The fix is mostly belt hardware — load limiters and pretensioners in the rear seats, the same parts the front has had since the nineties.
What it means for a used car
Every vehicle built before 2022 was designed against the old test. Its rating tells you what happened to the driver and nothing about the seat your children sit in.
That is not a reason to avoid older cars, and the back seat remains the correct place for a child seat at every age the manufacturer specifies. But if you are choosing between a 2021 and a 2024 of the same model, this is a real difference between them that does not appear in the star rating, the price, or the listing. Check the rating for the exact model year before assuming a strong safety reputation covers it.
Frequently asked questions
What changed in the updated moderate overlap test?
IIHS added a second dummy, the size of a small woman or a twelve-year-old, in the seat behind the driver. The crash itself is unchanged. Of 227 vehicles rated Good under the original version, only 58 kept that rating and 147 fell to Poor or Marginal.
Is the back seat less safe than the front?
In modern cars it can be, for adults. The front seat received decades of belt and airbag development — load limiters, pretensioners, occupant-specific airbags — and much of that never reached the rear. The updated test measures exactly that gap.
Should children still sit in the back?
Yes. The back seat remains the correct place for children at every age manufacturers specify, primarily because of front airbag risk. The finding is about rear belt hardware in adult-sized crash measurements, not a reason to move a child seat forward.
Denis Kataev is a serial solopreneur and the founder of ForCar. With 15+ years in software engineering and 10 years in SEO, he builds data-driven products end to end — backed by a sharp eye for design. At ForCar he mines proprietary vehicle datasets, turning raw numbers into buying advice you can actually trust.
Keep reading
All articles →
68,482 Owners Wrote That It Happened Without Warning. Most of Them Meant the Engine.
Most faults announce themselves. A noise, a light, a smell, something that gets worse over a week. A measurable share do not, and the complaint text says which systems behave that way.
25,641 Cars Caught Fire Without Crashing. Only 1,239 Burned After a Crash.
Car fires are filed under crash safety in most people's heads. In the complaint record they are almost entirely a separate phenomenon: the overwhelming majority start with no impact of any kind.
After a Death, the Complaint Takes 467 Days. After a Fire, 179.
We have measured the reporting delay by component before. Sorting the same million records by what actually happened to the people involved reverses the expected order completely.
Mark ForCar as a preferred source and our data shows up first in your Google results. One click, no account.