Nobody Crash-Tests the Third Row
Every three-row SUV and minivan on sale is advertised to families who will put children in the back. No US consumer crash test has ever placed a dummy there. This page sets out what is actually known.
Sources: IIHS and NHTSA published test protocols; NHTSA FARS person table, 2020β2024
Where the dummies actually sit
Both US rating programs stop at the second row:
- IIHS moderate overlap front (updated). An average-male dummy in the driver's seat and a second dummy the size of a small woman or 12-year-old, belted in the second row behind the driver.
- IIHS side (updated). Driver and second-row dummies on the struck side.
- NHTSA New Car Assessment Program. Driver and front passenger in the frontal test; driver and a rear-seat dummy in the side tests.
No dummy, in any of these tests, is seated in the third row. A vehicle can hold the highest award in the country without a single measurement ever being taken in the seats its advertising is aimed at.
The only evidence that exists
Since nothing is measured in a laboratory, the alternative is what happened in real crashes. FARS records a seating position for every occupant. Across 2020β2024, among people riding in vehicles involved in a crash that killed someone:
| Row | Occupants | Died | Share who died | 95% interval | Restrained |
|---|---|---|---|---|---|
| Front row | 355,632 | 149,278 | 42.0% | 41.8%β42.1% | 66% |
| Second row | 44,789 | 10,199 | 22.8% | 22.4%β23.2% | 64% |
| Third row | 2,812 | 354 | 12.6% | 11.4%β13.9% | 72% |
The third row shows the smallest share. Because third-row occupants are overwhelmingly children, and children are more fragile than adults in the same crash, it is worth checking whether age is doing the work. It is not β restricting to occupants under 16 gives the same ordering:
| Row | Children | Died | Share who died | 95% interval |
|---|---|---|---|---|
| Front row | 6,694 | 1,551 | 23.2% | 22.2%β24.2% |
| Second row | 20,104 | 2,853 | 14.2% | 13.7%β14.7% |
| Third row | 1,798 | 172 | 9.6% | 8.3%β11.0% |
What this does not prove
It would be easy, and wrong, to read the table as βthe third row is the safest seat.β Three things stand in the way.
- The denominator is a fatal crash, not a trip. Every vehicle counted here was in a crash that killed someone. A row far from the point of impact will show a lower share for that reason alone, and most fatal crashes are frontal.
- Impact direction is not controlled. In a rear impact the third row is the crumple zone, and this measure cannot separate that case out.
- Who rides where is not random. Third rows are occupied mainly in large SUVs and minivans, which differ from the overall fleet, and mainly on family trips rather than the late-night and high-speed driving that fills the fatal-crash record.
IIHS's own research runs the other way on the second row: they added the rear dummy to the moderate overlap test after finding that in newer vehicles, belted second-row occupants face a higher risk of fatal injury than those in front. Not because the back seat got worse, but because front seats gained better airbags and advanced belts that the back rarely receives. The same logic should make anyone cautious about the third row, which gets less engineering attention still.
The honest summary
The third row is not the death trap it is sometimes assumed to be β the outcome data gives no support for that. But βnot a death trapβ is the most that can be said. There is no measurement, no rating and no way to compare one vehicle's third row against another's. If that matters to you, it is worth telling IIHS and NHTSA so, because the gap will close only when someone asks for it.
What you can compare is everything forward of it.