We have already explained airbags and passive safety and the materials a body structure is built from — the hardware. What we haven’t explained is how anyone actually measures whether that hardware works, and what the star rating stamped on every new car’s brochure actually represents. Euro NCAP (the European New Car Assessment Programme) is not an EU regulator — it’s an independent, non-profit consortium of European governments, motoring organisations and safety research bodies — and its rating is not a pass/fail exam but the weighted result of four separate test categories that get demonstrably harder to pass every single year.
Four categories, one weighted score
A car’s overall star rating is a weighted composite of scores in four areas, not a single crash test. Adult Occupant Protection measures how well the structure and restraint systems (the belts and airbags covered in our earlier article) protect the driver and front passenger across frontal, side and whiplash tests. Child Occupant Protection repeats a version of that assessment specifically for child restraint systems fitted in the rear. Vulnerable Road User Protection tests how the car’s front end and active safety systems perform against pedestrians and cyclists, rather than against other vehicle occupants. Safety Assist scores the presence and real-world performance of active systems — autonomous emergency braking, lane support, speed assistance — the ADAS features we’ve already detailed elsewhere. A car needs to perform well across all four, not just the traditional frontal-crash category most people picture, to earn five stars.
The stars are a moving target on purpose
The single most misunderstood fact about Euro NCAP is that its protocol is revised on a multi-year roadmap, with the passing bar raised each cycle — which is why a five-star rating from 2015 does not mean the same thing as a five-star rating from 2025 tested against a harder, more comprehensive protocol. This is deliberate: Euro NCAP’s own stated purpose is to keep pushing manufacturers ahead of where regulation requires them to be, not to hand out an easy, stable grade. Where do the new test scenarios that keep raising that bar actually come from? A large share traces directly back to EU-funded research years earlier. PROSPECT (2015-2018, €6.93M, coordinated by IDIADA in Spain) developed and validated the expanded scenarios and improved detection performance behind autonomous emergency braking for pedestrians and cyclists (AEB-VRU) — test criteria that Euro NCAP subsequently folded directly into its Vulnerable Road User Protection category. A test you can find in this year’s protocol quite plausibly started as a research scenario defined years earlier in a project like this one.
The dummy problem, and the shift to virtual testing
A physical crash test has an unavoidable limitation: it can only ever test as many body types as there are dummies to put in the car, and for decades the industry standard dummy represented one reference build — a limitation with real consequences, since real occupants range enormously in age, size and physiology. VIRTUAL (2018-2022, €7.00M, coordinated by VTI, the Swedish national road and transport research institute) tackled this directly by building open-access virtual testing tools and scalable Human Body Models, covering different ages and sizes of both car occupants and vulnerable road users, and validating computer simulation as a genuinely reliable substitute for at least some physical crash tests. The direction of travel this points to is significant: future safety assessment doesn’t have to be limited to whatever a handful of physical dummies can represent, but can in principle be tested against body types a physical lab could never afford to stock.
Why it matters beyond the badge on the brochure
None of this makes a star rating meaningless — it makes it a snapshot against a specific, dated, and genuinely demanding protocol, not an absolute, timeless measure of safety. Reading a rating well means checking the year it was tested against, not just the number of stars, and understanding that behind every new test category sits years of research, often EU-funded, quietly deciding what “safe” is going to mean next.
Photo: © Car-Shooters