The front of a car is built to absorb a crash it meets squarely. A crash that catches only the corner, a tree, a utility pole, the corner of another car, can go around those parts and into the cabin where the people are.
The test. The Insurance Institute for Highway Safety describes it this way: “The small overlap front test represents a crash in which the front corner of a vehicle collides with another vehicle or an object like a tree or utility pole.” In the test, “A vehicle travels at 40 mph into a rigid barrier with 25% of the vehicle’s front overlapping.” The Institute added the driver-side version to its crash tests in 2012, and today a vehicle is tested “once with the overlap on the driver side and once on the passenger side” (IIHS; IIHS, 2023).
Why the corner is hard. “Small overlap crashes pose a challenge because they bypass a typical vehicle’s main front structure” (IIHS, 2016). In this kind of test, the Institute says, “there is often no direct impact with the vehicle’s frame rail and therefore the occupant compartment and other structures must manage the bulk of the energy of the crash” (IIHS, 2023). The parts designed to fold and buy time, the ones Crush and Ride-Down describe, may barely be touched. The energy has to go somewhere, and it goes into the structure around the people.
What happens inside. The Institute measures “intrusion into the occupant compartment at key locations,” and its test dummy carries “Sensors in the head, neck, chest, legs and feet” to record injury risk (IIHS). In a 2016 study of the passenger side, it tested seven small SUVs that had good ratings on the driver side, and “Only one” of them performed at a level corresponding to a good rating on the passenger side. In one vehicle the door hinge pillar “tore off completely,” and in another the door opened. In the vehicle whose door opened, the most intrusion measured on the passenger side was 13 inches more than in the same model’s driver-side test (IIHS, 2016). Inside the car, the body then meets whatever the cabin has become, which is the subject of The Third Collision.
On the road. In 2023 the Institute checked its ratings against real crashes. “A driver in a good-rated vehicle is 12 percent less likely than a driver in a poor-rated one to be killed in a frontal crash.” An acceptable rating “was associated with an 11 percent lower risk of dying,” and a marginal rating “with a 5 percent lower risk,” both against poor. The study counted frontal crashes of every kind, not only small overlap crashes, and the Institute says “the effect of a good rating is likely much larger among those specific crashes” (IIHS, 2023).
One side is not the other. The 2016 release opened with the plainest finding: “good protection doesn’t always extend across the front seat” (IIHS, 2016). A good rating earned with the overlap on the driver’s side did not, by itself, promise the same protection for the person in the other front seat.
What the car can prove. The vehicle is evidence. Where it was struck, how much of its front was engaged, how far the cabin moved inward, whether a door opened: those are measured on the car itself, and they are gone once it is repaired or crushed for scrap (Spoliation). Whether an injury came from the crash, or from how the vehicle held up in it, are two different questions, and the second is Crashworthiness. The Fixed Object explains why a narrow object is so hard on a car. This page is about where on the car the blow lands. Closing Speed explains how hard it lands.
Crash protection is not one number for the whole car. It depends on where the blow lands, and a glancing blow can land where the car has the least to give.
Insurance Institute for Highway Safety, Small overlap front test; IIHS, Small overlap front crash rating program delivers real-world benefits (March 2, 2023); IIHS, Small overlap gap: vehicles with good driver protection may leave passengers at risk (June 23, 2016). This page names no vehicle model; the 2016 passenger-side ratings were provisional. This is general information, not medical or legal advice. Text and figures as published at the time of this publication.