Informed for Life

Step 1 — Start With IIHS Awards

Every vehicle on our list must first earn the IIHS Top Safety Pick+ (TSP+) award, the highest recognition from the Insurance Institute for Highway Safety. This is our baseline filter — only vehicles that have already passed IIHS's rigorous evaluation are considered.

Step 2 — Evaluate Individual Crash Test Ratings

We then look at the vehicle's specific crash test scores. IIHS rates each test as Good, Acceptable, Marginal, or Poor. The tests we evaluate are:

Step 3 — Factor In Real-World Death Rates

Crash tests are conducted in controlled labs. To go beyond lab results, we incorporate demographic-adjusted driver death rates — real-world data showing how many drivers actually die per million registered vehicle years, broken down by vehicle class.

We use the vehicle's class death rate (e.g., "Midsize SUVs" or "Large luxury SUVs") and compare it to the national average of 38 deaths per million registered vehicle years.

Why class-based rates? Individual model death rates require years of data to be statistically reliable. Class-based rates (e.g., "Midsize luxury SUVs") are more stable and still capture the safety advantage of vehicle size and weight. Source: IIHS Status Report / HLDI research.

The Safety Score (0–100)

Every TSP+ vehicle receives a Safety Score combining crash performance and real-world outcomes:

Tier Criteria

How We Pick the Top 5 and #1 Safest

Beyond the tier system, we apply four additional filters to identify the absolute safest vehicles. These criteria go beyond crash test ratings to consider factors that affect real-world survival:

See our full analysis: The #1 Safest Car →

How We Sort Within Each Tier

Within each tier, vehicles are sorted by Safety Score (highest first). A higher Safety Score means better crash test results and a lower class death rate. If two vehicles share the same Safety Score, they are sorted alphabetically by make and model.

Our Data Sources

We update our rankings whenever IIHS publishes new Top Safety Pick awards or death rate data. Our code is deterministic — the same inputs always produce the same rankings.

Limitations & Further Reading

No safety ranking system is perfect. Here are the key limitations of our approach:

• Death rates partly reflect driver behavior, not just vehicle design. Minivans have low death rates partly because parents drive cautiously. Sports cars have high death rates partly because they attract aggressive drivers. IIHS adjusts for driver age and sex, which removes a significant portion of this bias — but cannot fully eliminate all behavioral factors.
• Class-level death rates mask individual model differences. Two midsize SUVs may have very different crash structures, but both inherit the same class death rate in our formula. For model-specific death rates, see our Used Car Safety Rankings.
• Crash tests cover specific scenarios. A "Good" rating in a 40 mph frontal offset test doesn't guarantee the same performance at 60 mph or in a rollover. Real-world crashes are far more varied than lab conditions.

Academic research: Economists Ben Shiller and Bhoomija Ranjan at Brandeis University have published research showing that coarse safety ratings (Good/Acceptable/etc.) hide meaningful differences between vehicles — and that continuous ratings could prevent ~1,850 US fatalities per year. Their site RealSafeCars.com takes a complementary approach, using machine learning to correct for population bias in death rates. See: "Are Coarse Ratings Fine? Applications to Crashworthiness Ratings" (Brandeis Working Paper, 2020).