When a car earns 5 stars from NHTSA, most shoppers treat it as a binary: safe or not safe. The reality is more useful than that. Within the 5-star tier, vehicles produce different forces on the crash dummy — and those differences are invisible in the star label. SafeScore translates those underlying measurements into a number that shows how much margin the dummy actually had. This article explains what that means in plain English.
NHTSA's star rating system condenses multiple sensor measurements from a 35 mph frontal crash test into a single tier from 1 to 5. A 5-star frontal rating means the vehicle kept measurements for head, chest, and neck injuries below federal thresholds with enough margin to qualify for the top tier. It is a meaningful achievement — a vehicle with 4 or fewer stars failed to keep those measurements as far below the thresholds.
But the star system is a category, not a measurement. Two vehicles that both earn 5 stars can have different underlying sensor readings. One may have a head injury measurement (HIC15) of 350 — comfortably below the federal limit of 700. Another may have a reading of 630 — still within the 5-star range, but considerably closer to the limit. Both are labeled 5 stars. The distance from the threshold is invisible in the star alone.
In NHTSA's full-frontal barrier test, an instrumented crash dummy sits in the driver and right-front passenger seats. Dozens of sensors record forces throughout the impact. SafeScore is built on three of those measurements:
SafeScore converts each of the three measurements into a percentage of its federal threshold, averages the three percentages, and subtracts from 100. A vehicle using 20% of the average threshold on average scores 80. A vehicle using 45% of the average threshold scores 55. Both may carry a 5-star NHTSA rating; SafeScore shows the difference underneath.
Formula: SafeScore = 100 − mean( HIC15÷700×100, chest_mm÷63×100, NIJ÷1.0×100 ). Equal weighting is used because each measurement is already normalized against its own injury threshold — no invented medical weights.
When a test includes both driver and passenger dummies, SafeScore uses the lower of the two scores — never the average. A car should protect both seats, not just the better one.
SafeScore is specifically a frontal crash injury-margin score. It does not incorporate NHTSA side or rollover ratings, IIHS test results, real-world crash outcomes, or driver behavior. NHTSA star ratings cover more configurations than SafeScore — side and rollover tests are evaluated separately and shown alongside SafeScore on each vehicle page. For a complete picture, use SafeScore as the resolution layer within NHTSA frontal results, alongside NHTSA side/rollover stars and IIHS ratings.
SafeScore matters most when you are deciding between two specific vehicles that both earn strong star ratings. If one vehicle scores 78 and another scores 61, the 78 left substantially more injury margin in the frontal crash test. That gap may be more useful than the fact that both earned 5 stars. The star rating tells you both passed; SafeScore tells you by how much.
It matters less when comparing a 5-star vehicle to a 3-star vehicle — the star gap already tells you something important. The star rating is a good first filter. SafeScore is the tool for comparing within the top tier.
Independence: SafeScore is calculated by SafeCarCompare from NHTSA public-domain crash test data. It is not issued or endorsed by NHTSA. All underlying measurements can be verified at api.nhtsa.gov.
SafeCarCompare shows injury margins from NHTSA crash-test data — beyond star ratings. Enter any two vehicles to see head, chest, and neck injury margins side by side.