← The Crash Report
The Gap

IIHS Put a Dummy in the Back Seat for the First Time. Only 2 of 16 Cars Survived.

For thirty years, U.S. crash tests measured what happens to the driver. Nobody in the testing lab bothered to check on the passenger sitting behind them. IIHS finally changed that with its updated moderate overlap frontal test, dropping a Hybrid III dummy into the second row for the first time. When they ran 16 new 2026 model year vehicles through the protocol, exactly two earned Top Safety Pick+.[1]

2 / 16
Vehicles earning TSP+ under the new rear-seat crash test protocol

Those two survivors: the Hyundai Palisade and the Toyota Prius. Every other vehicle tested fell short of the top award, many specifically because of what the rear-seat dummy endured. IIHS Senior Research Engineer Marcy Edwards, who developed the new evaluation, called it "the first frontal crash test in the U.S. to include a rear-occupant dummy."[2] That sentence should disturb you. Crash testing has existed since 1995. Children have been riding in back seats the entire time. Nobody was checking.

Consider what the new test measures. Sensors in the second-row dummy record injury risk to the head, neck, chest, abdomen, and thigh. Video cameras and greasepaint confirm whether the dummy's head hits the vehicle interior or gets dangerously close to the front seatback. A pressure sensor monitors shoulder belt position on the torso. And the test watches for "submarining," where the occupant slides forward beneath the lap belt into an abdominal injury configuration. Less than half of new vehicles sold today have advanced restraint systems in the rear seat.[2]

FARS fatal crash data from 2014 to 2023 has been screaming about this gap for a decade. Across all vehicle classes in the dataset, sedans show a crash lethality ratio of 0.643, meaning in nearly two-thirds of fatal crashes, the vehicle's own occupant dies.[3] That number includes rear-seat passengers who had no front-seat airbag, no pretensioner, no force limiter. Front-seat crash technology improved year after year. Back-seat technology mostly stayed the same. IIHS's own prior research found that rear-seat occupants faced substantially higher injury risks than front-seat occupants in frontal crashes, even in vehicles rated "good" under the old test protocol.

Some of the specific failures in the new batch are illuminating. Cadillac's CT5 earned poor ratings in two key crash tests, which is remarkable for a vehicle starting above $40,000.[4] Cadillac sedans in FARS carry an average fatality rate of 2.60 per 100 million vehicle miles traveled, the highest among luxury marques in the dataset. Paying more for a Cadillac badge has not historically purchased better crash outcomes, and now the lab data is confirming what the roads already showed. Audi's A3, Lexus IS, Nissan Kicks, and the Toyota Tacoma crew cab each earned marginal in a key test. Consumer Reports downgraded the A3's safety verdict from "Better" to "Basic."

Then there is the Palisade paradox. Hyundai's large SUV earned top marks in the toughest crash test available. It also triggered a 61,093-vehicle recall in March 2026 after a two-year-old girl in Ohio was killed when the power seat did not detect contact with her body.[5] NHTSA found that the second- and third-row power seats "may not detect contact with an occupant or object as intended." A vehicle that protects rear-seat occupants from a 40 mph frontal crash killed a child in its own cabin at zero miles per hour. Crash tests measure crash energy. They do not measure the cumulative engineering decisions a manufacturer makes about every other component in the car.

What to do with this: If you are buying a new car and anyone ever sits in the back seat, check whether it earned good ratings under IIHS's updated moderate overlap test. Vehicles rated before the protocol change are being re-evaluated. Front-seat safety ratings are largely irrelevant to passengers in the second row. Look specifically for pretensioners and force limiters in the rear seat belts, which are the same technologies that revolutionized front-seat survival. If the automaker does not list them in the spec sheet, they probably are not there. The Prius earned TSP+ starting under $30,000. The CT5 failed it at $42,000. Price is not protection.

Sources & References

  1. Autoblog, “IIHS Tested 16 New Cars, But Only Two Earned The Top Safety Award,” May 15, 2026. autoblog.com
  2. IIHS, “New crash test spotlights lagging protection for rear passengers.” iihs.org
  3. NHTSA, Fatality Analysis Reporting System (FARS), 2014–2023. nhtsa.gov
  4. Consumer Reports, “Safest New Cars of 2026, According to the IIHS.” consumerreports.org
  5. Reuters, “Hyundai to recall over 61,000 Palisade SUVs in US after fatal incident, NHTSA says,” March 20, 2026. reuters.com

Source: NHTSA FARS 2014–2023; IIHS 2026 model year test results. Crash lethality ratio calculated as occupant deaths per fatal crash event. See methodology for caveats.