YearToAvoid

Data Sources

Everything on this site is computed from public U.S. government datasets. Here is exactly what we use and where it comes from.

NHTSA Vehicle Owner Complaints

All safety-related defect complaints received by NHTSA’s Office of Defects Investigation, published as daily-updated flat files. We use complaints received from 2010 onward for model years 2010–2025.

Fields used: make, model, model year, component, crash, fire, injuries, deaths, incident date, filing date, mileage, narrative.

https://www.nhtsa.gov/nhtsa-datasets-and-apis · flat files

NHTSA Safety Recalls

Every Part 573 safety recall campaign since 2010, one row per campaign / make / model / year.

Fields used: campaign number, component, defect summary, consequence, remedy, Do Not Drive and Park Outside advisories, report date.

https://www.nhtsa.gov/nhtsa-datasets-and-apis · flat files

NHTSA vPIC (Product Information Catalog)

Canonical make and model names by model year, and the VIN decoder used by our recall check tool.

https://vpic.nhtsa.dot.gov/api/

NHTSA Safety Ratings (NCAP)

5-Star Safety Ratings overall score, when NHTSA has crash-tested that model year.

https://www.nhtsa.gov/ratings

EPA Fuel Economy (fueleconomy.gov)

Official city, highway and combined MPG plus annual fuel cost for every configuration sold in the U.S.

https://www.fueleconomy.gov/feg/ws/ · flat files

Update cadence

The database is rebuilt weekly from the latest flat files. The current build includes complaints filed through 2026-09-24. Pages show the data date in their footer.

Known limitations

  • Complaints are unverified consumer reports. NHTSA does not confirm each one.
  • No sales data exists in these sources, so complaint counts are not normalized per vehicle sold. We only compare years of the same model.
  • NHTSA sometimes lists trims or hybrids as separate models (for example “CR-V” and “CR-V Hybrid”). We keep NHTSA’s split rather than guess.
  • Mileage at failure is present in only part of the complaints.