The sound hits first. As loud as a gunshot, it fills the entire hangar-sized space in an instant. You can almost feel the sonic reverberations from the impact, almost smell the kinetic ripple through the air. And just like that, it's over. The measurable impact lasts barely two tenths of a second.
It's rare to witness a crash with this kind of force in real time. When it happens in the real world, you're almost never prepared—you catch it on the periphery, already in progress. At the Insurance Institute for Highway Safety's Vehicle Research Center in Ruckersville, Virginia, the experience was something else entirely. Hollywood-grade lighting illuminates the action. Cameras record the trauma at up to 500 frames per second. Our attention was focused, our eyes trained—and even still, we were hardly prepared for the shock.
The subject was a brand-new 2026 Mazda CX-5. Mazda had invited members of the media to observe a moderate overlap front crash test—the kind that sends a vehicle into a deformable barrier at 40 mph, with the barrier covering 40 percent of the front end on the driver's side. When it was over, the front driver's side was crushed in, but only to the leading edge of the A-pillar. From there back, the cabin was largely preserved. Every airbag had deployed. Paint markings on each one showed exactly where the crash dummy's head had made contact. The car looked destroyed. But in a real-world scenario, the occupants would have a real chance to walk away.
That chance is exactly what crash testing is designed to measure. Understanding it makes you a better shopper.
- What Is a Crash Test?
- How Does IIHS Differ from NHTSA?
- What Does the IIHS Test?
- What Do IIHS Crash Test Ratings Mean for Shoppers?
- How Cars Have Improved Over Time
- What's Next for Vehicle Safety
- Frequently Asked Questions
What Is a Crash Test?
Think of a crash test like a medical exam. It's less about the image itself and more about what it tells us. Although the IIHS provides ratings and accolades, the exercise of crashing cars is ultimately an analysis designed to estimate risk. Specifically, crash tests estimate the injury risk for real people in a collision.
It's also important to note that the crash test dummies inside the vehicle aren't a stand-in for any particular type of person. They are calibrated instruments measuring accelerations, forces, and the movement of body parts relative to one another. Those measurements are then used to calculate the probability of specific injuries in a real-world crash.
Although the IIHS provides ratings and accolades, the exercise of crashing cars is ultimately an analysis designed to estimate risk. Specifically, crash tests estimate the injury risk for real people in a collision.
IIHS tests with both a mid-sized male dummy and a small female dummy, although those names don't describe the only populations they protect. Jessica Jermakian, SVP of Vehicle Research for IIHS, was straightforward about this when we spoke at the facility. "The crash test dummies are tools," she said. "They're not real people. They shouldn't be treated like people."
The facility itself reinforces how seriously this work is taken. The lighting is production-grade and the cameras record at an incredibly high 500 frames per second, allowing for detailed slow-motion capture.
500 fps is fast enough to produce crystal clear slow-motion footage of a water balloon popping—imagine seeing the rubber peel away before the water inside collapses from gravity. The footage serves two purposes: precise frame-by-frame engineering analysis, and public communication. IIHS publishes its crash test videos because consumer awareness drives manufacturer behavior. When buyers understand what these tests show, automakers have more incentive to perform well in them. It's accountability through transparency, and it works.
How Does IIHS Differ from NHTSA?
Shoppers who have spent time looking at window stickers have seen crash test ratings from the National Highway Traffic Safety Administration (NHTSA). These ratings are different from the IIHS' evaluation, and critically, the two organizations are also not duplicating each other's work.
NHTSA is a federal regulatory agency. Its crash testing program sets minimum safety standards and provides consumer information on a five-star scale. The IIHS, on the other hand, is an independent nonprofit funded by auto insurers and operating a four-tier system: Good, Acceptable, Marginal, and Poor.
The key distinction is that IIHS is designed to fill gaps in the regulatory environment—to test things NHTSA isn't testing and to raise the bar beyond what federal standards require. Jermakian offered a concrete example: the moderate overlap test we witnessed, which involves a rear-seated occupant, is the first frontal crash program in the United States to evaluate rear occupant protection. It's not part of NHTSA's regulatory program. It's not part of NHTSA's consumer information testing. IIHS identified the gap and built a test to address it.
"Once there is a regulation that makes our testing obsolete, we'll sunset that program."
This dynamic also works in reverse. When a federal regulation catches up to what IIHS is testing, IIHS retires the program. "Once there is a regulation that makes our testing obsolete, we'll sunset that program," Jermakian explained, citing the roof crush program as an example. Once federal standards advanced to reflect the current science, IIHS ended its own evaluation—because continuing it would no longer give consumers any additional useful information.
If you're focused on safety while shopping for a car, the practical guidance is simple: use both IIHS and NHTSA ratings. They answer different questions.
| IIHS | NHTSA | |
|---|---|---|
| Type of organization | Independent nonprofit | Federal regulatory agency |
| Funded by | Auto insurers | U.S. government |
| Rating scale | Good / Acceptable / Marginal / Poor | 1–5 stars |
| Key designation | Top Safety Pick / Top Safety Pick+ | Overall Vehicle Score |
What Does the IIHS Test?
The moderate overlap front crash test is one of several evaluations the IIHS administers. Rather than striking the center of the vehicle, the barrier hits 40 percent of the front end on the driver's side, stressing the structures most likely to fail in a real-world offset collision: the A-pillar, door frame, and footwell.
Other categories include a small overlap front test (just 25 percent of the front end), a passenger-side small overlap test, a side test that simulates being struck by a larger vehicle, a whiplash prevention evaluation, rear crash prevention, and front crash prevention evaluations covering vehicles, pedestrians, and increasingly, bicyclists.
Headlight performance is also evaluated—a category that has quietly driven significant industry improvement over the past decade and one that affects Top Safety Pick (TSP) and Top Safety Pick+ (TSP+) eligibility. The IIHS also evaluates safeguards for partial driving automation, a newer category reflecting how quickly hands-free driving technology has spread across the market.
It's also worth noting that automakers don't submit their cars for evaluation completely cold. "The automakers, generally speaking, run our tests in-house before we ever run our tests," Jermakian told us. "So they have a good idea of how their vehicle is going to perform before they ever bring it." Mazda's decision to invite journalists to this particular test was a statement of confidence.
| Test | What it evaluates |
|---|---|
| Moderate overlap front | A 40 mph impact covering 40% of the front end on the driver's side — the A-pillar, door frame, and footwell |
| Small overlap front | A 40 mph impact covering just 25% of the front end, conducted separately on the driver's side and the passenger's side |
| Side | A moving barrier simulating a larger vehicle striking the driver's side door |
| Whiplash prevention | Evaluates front seat and head restraint design for rear-impact neck injury risk |
| Front crash prevention: pedestrian | Evaluates automatic emergency braking for pedestrians and bicyclists |
| Front crash prevention: vehicle-to-vehicle 2.0 | Evaluates automatic emergency braking in vehicle-to-vehicle crash scenarios |
| Headlights | Rates low and high beam visibility |
What Do IIHS Crash Test Ratings Mean for Shoppers?
The best rating a vehicle can earn in any particular test is Good. That said, IIHS evaluates vehicles across multiple test types, and a strong result in one area doesn't guarantee strong results across the board. The full scorecard matters, not just the headline designation.
Vehicles that meet the minimum thresholds across all evaluated categories earn a Top Safety Pick or Top Safety Pick+ designation. The difference between the two comes down to front crash prevention performance—specifically, TSP+ requires a Good rating in pedestrian automatic emergency braking (AEB), versus Acceptable or Good for TSP. Acceptable or Good performance in vehicle-to-vehicle front crash prevention 2.0 is another requirement for a TSP+ designation, but it isn't required for TSP at all.
Headlights have tripped up plenty of otherwise strong performers, too—both designations require Acceptable or Good headlights, so a car can be a vault from the front and still miss the award because its lights are mediocre.
For most testing, the IIHS purchases vehicles through normal retail channels, the same way a consumer would. When a manufacturer wants to supply a vehicle for testing, it contributes the cash and the IIHS purchases the vehicle itself.
| Requirement | Top Safety Pick | Top Safety Pick+ |
|---|---|---|
| Small overlap front | Good | Good |
| Moderate overlap front | Good | Good |
| Side | Good | Good |
| Headlights (standard) | Acceptable or Good | Acceptable or Good |
| Pedestrian front crash prevention (standard) | Acceptable or Good | Good |
| Vehicle-to-vehicle front crash prevention 2.0 (standard) | Not required | Acceptable or Good |
How Cars Have Improved Over Time
The 2026 CX-5 uses what Mazda calls a Ring Structure, a unibody architecture engineered to distribute crash energy around the occupant compartment rather than through it. The post-crash car we saw bore this out. The front driver's side absorbed the impact. The A-pillar held. Behind it, the cabin was intact.
In the IIHS facility's front hall, visitors and staff alike are treated to a clear demonstration of progress. A 1959 Chevrolet Bel Air and a 2009 Chevrolet Malibu—two vehicles IIHS brought together for a crash test staged as part of its 50th anniversary—demonstrate how much crashworthiness had improved over half a century. In the Malibu, the occupant compartment remained intact. The front end was destroyed; the cabin was not. In the Bel Air, the occupant compartment collapsed entirely. Survival didn't just look unlikely—a rescue team would have needed to spend precious additional time cutting the victim out of the mangled metal that had intruded into the cabin. Standing next to both vehicles, the contrast isn't subtle. It doesn't require an engineering degree to interpret.
Similarly, adjacent to the IIHS facility's crash-testing area is a large room filled with crumpled vehicles. It's a strange museum of sorts—a collection of post-test vehicles in various states of destruction, spanning several decades. Walking through it conveys a visual story that statistics alone can't match.
A 2000s Ford Focus with its driver's-side doors completely crushed inward. A pair of Chevrolet Blazers from different eras, showing more recent progress in a compressed timeframe. A first-generation Volkswagen Tiguan and a third-generation Kia Sportage, showing how two cars of similar age can perform wildly differently in the roof strength test that the IIHS has since retired—a program that drove real improvements before federal standards caught up, and a good illustration of how the organization approaches its own obsolescence. (Head's up: you were much, much safer in the Volkswagen.)
What's Next for Vehicle Safety
Traffic fatalities in the United States spiked sharply during the pandemic years, climbing from 36,355 in 2019 to 43,230 in 2021. They've declined since—40,901 died on U.S. roads in 2023—but that number remains more than 8,000 deaths above the modern low set in 2014. Crash test performance has never been better. The death toll tells a different story.
IIHS launched its 30 by 30 initiative in response. The goal is to reduce U.S. traffic fatalities 30 percent by 2030, benchmarked against 2022 data. The 2022 figure is used because crash fatality data lags by roughly two years. Hitting the target would return annual fatalities to their lowest point in more than a decade.
The reason for the gap between vehicle improvements and fatality trends comes down to behavior. "Half of the fatalities are people who are unbelted," Jermakian told us, "which is astounding, considering our belt use rate is over 90 percent in this country." The remaining deaths cluster around speeding, distraction, and impairment—problems that structural engineering can't fully address. You can build a better cage, but you can't build your way out of someone choosing not to wear a seatbelt.
So IIHS is expanding its focus. A new category of evaluation is in development around vehicle technologies that encourage safe driving behavior—systems designed to address the belt, speed, and attention problems that crashworthiness advances can't touch. Front crash prevention that detects bicyclists is also coming. These aren't replacements for crash testing. They're parallel tracks, each targeting a different slice of the fatality picture.
Frequently Asked Questions (FAQ)
What is the difference between IIHS and NHTSA ratings?
The IIHS is an independent nonprofit funded by auto insurers. NHTSA is a federal regulatory agency. Both conduct crash testing, but they aren't covering the same ground—IIHS fills gaps in the regulatory environment, testing things NHTSA isn't evaluating. A strong score from one doesn't substitute for a strong score from the other. Use both.
What does a "Good" IIHS rating mean?
Good is the top rating on IIHS's four-tier scale: Good, Acceptable, Marginal, and Poor. It means the vehicle performed at the highest level in that specific test. It does not mean the vehicle has been rated Good across all evaluations—check the full scorecard before drawing conclusions.
What is the difference between Top Safety Pick and Top Safety Pick+?
Both require Good ratings in the small overlap front, moderate overlap front, and side tests, plus Acceptable or Good headlights. The difference comes down to front crash prevention. TSP+ requires a Good rating in pedestrian AEB, versus Acceptable or Good for TSP, and adds a requirement for Acceptable or Good performance in vehicle-to-vehicle front crash prevention 2.0 that TSP doesn't have.
Are crash test dummies representative of real people?
No—and that's intentional. Crash test dummies are engineering tools that measure accelerations, forces, and body part movements. Researchers translate those measurements into injury risk estimates for real people. IIHS uses both a mid-sized male dummy and a small female dummy, and improvements driven by either benefit all occupants. The names are engineering classifications, not descriptions of who the dummies exclusively protect.
Do automakers know how their cars will perform before IIHS tests them?
Generally, yes. Most automakers run IIHS protocols in-house before submitting a vehicle, so they arrive with a reasonable expectation of results. IIHS is also transparent in advance about what it's testing and what vehicles will be evaluated on. A manufacturer inviting journalists to watch is typically a sign they like their odds.
What is the moderate overlap front crash test?
The moderate overlap front crash test sends a vehicle into a deformable barrier at 40 mph, with the barrier covering 40 percent of the front end on the driver's side. It's designed to stress the structures most likely to fail in a real-world offset collision—the A-pillar, door frame, and footwell—rather than the center of the car, where structural integrity is typically strongest.
Have cars gotten safer over time?
Dramatically. In 2009, IIHS staged a direct demonstration by crashing a 1959 Chevrolet Bel Air into a 2009 Chevrolet Malibu. In the Malibu, the occupant compartment remained intact despite severe front-end damage. In the Bel Air, it collapsed entirely. The institute's then-president called the difference "night and day." The post-test vehicles tell the story more vividly than any statistic.
What is the IIHS 30 by 30 initiative?
The 30 by 30 initiative is IIHS's goal to reduce U.S. traffic fatalities 30 percent by 2030, measured against 2022 data. Reaching the target would return annual fatality numbers to their lowest point in over a decade. The initiative focuses on behavioral factors—belt use, speeding, distraction, impairment—which account for a disproportionate share of deaths that structural improvements alone can't prevent.
Should I use IIHS or NHTSA ratings when buying a car?
Both. They test different things and neither is a substitute for the other. A vehicle that performs well in one program isn't guaranteed to perform well in the other. Use both ratings as part of your research, and look at the full scorecard for each—not just the summary designation.
