Vehicle-year public record
Compare NHTSA complaint reports, recall campaigns and manufacturer communications for a specific make, model and year. Counts describe records in the snapshot, not failure rates.
Choose a vehicleCar research lab
Research a car before you price the warranty.
Join model-year safety records, vehicle configuration facts and repair-cost scenarios, then test a real contract against the same evidence. Each layer keeps its source date and its limits.
339,560complaint/model-year records
2,742vehicle years
11,636EPA configurations
2,931public issue-file paths
Original studies · September 2026
Findings you can check, not just headlines
Three reproducible investigations built from the site's data. Each includes sources, downloadable rows, an original chart and the limits of the conclusion.
Cost-model experiment$524.32modeled difference for eight labor hours
The same job, a different city
An eight-hour labor scenario differs by $524.32 between the lowest and highest modeled metros. See all inputs, the calculation and its limits.
Read the study and get the data →Public-record audit155 / 177zero-complaint model-years with other recordsWhat zero complaints leaves out
155 of 177 model-years with zero complaints still have recall or manufacturer-communication records. Explore the snapshot and its missing denominators.
Read the study and get the data →Source-coverage audit134model-years without a matched EPA configurationWhen the model name is not enough
134 model-years lack an EPA configuration match. See where trim ambiguity and incomplete source joins limit a warranty estimate.
Read the study and get the data →Three questions, one vehicle
What the research can answer
The useful result is not a generic reliability score. It is a traceable view of what has been reported, what repair scenarios fit the vehicle and where a quoted contract would break even.
Configuration-aware repair exposure
EPA configuration matches help remove clearly incompatible repair scenarios. The remaining costs are planning ranges, not predictions that a component will fail.
Review repair researchContract break-even math
Enter a real plan price and deductible to find the repair bill needed for one covered claim to recover that amount. Coverage terms still control the outcome.
Run the calculatorSource and date register
Four layers, kept separate
Observed records, configuration facts, cost models and external research links answer different questions. Combining them does not turn any one layer into a failure forecast.
339,560 complaint/model-year records, 11,020 recall/model-year records and 864,817 communication/model-year records. IDs are deduplicated within each model-year; these sums are not globally unique reports, campaigns or documents.
Snapshot or source date: 2026-09-03
11,636 EPA configurations cover 2,608 model-year records. Fuel type, drive layout and equipment flags narrow the repair scenarios without borrowing facts from another year.
Snapshot or source date: 2026-08-07
The repair layer covers 44 major scenarios. RepairPal is the published benchmark source where a matched benchmark exists; local labor assumptions then create broad planning ranges.
Snapshot or source date: 2026-06-18
2,359 vehicle-year paths and 2,931 issue-file paths help readers continue their research. Public factual metadata adds explicit covered years, evidence labels and citations for 1,221 canonical issue records. Article prose, images and dynamic owner counts are not imported.
Snapshot or source date: 2026-09-03
Evidence-rich starting points
See how the records join on a real car
These examples have both model-year evidence and enough compatible repair scenarios for a useful profile. They are not a best, worst or reliability ranking. Complaint totals are not adjusted for sales, mileage, severity or reporting behavior.
2019
Toyota RAV4
897 complaint reports · 8 recall campaigns
Open model-year research →2018Ford F-150
2,285 complaint reports · 19 recall campaigns
Open model-year research →2018Honda CR-V
2,127 complaint reports · 5 recall campaigns
Open model-year research →2018Tesla Model 3
1,009 complaint reports · 19 recall campaigns
Open model-year research →2019Subaru Outback
1,032 complaint reports · 3 recall campaigns
Open model-year research →2015Jeep Grand Cherokee
1,355 complaint reports · 10 recall campaigns
Open model-year research →How new data enters the lab
New snapshots can rebuild the research without replacing the site.
Bulk NHTSA and EPA files are normalized into the existing model-year records. Repair benchmarks and the public FixCatalogue sitemap keep separate dates. A refreshed source can change counts and comparisons, so each result remains tied to its snapshot.
Limits that matter
This is decision support, not a reliability verdict.
Complaint reports are submitted records, not failures per vehicle sold. Recall campaigns do not reveal how many vehicles remain unrepaired. Manufacturer communications cover many kinds of service information.
Repair ranges are planning estimates, not live shop quotes or predictions that a repair will occur. Only the signed contract establishes exclusions, waiting periods, deductibles, labor allowances and payout limits.
FixCatalogue contributes public sitemap paths and attributed factual metadata, including covered years and official source citations. AutoWarranty.Autos does not copy its article text, images, owner submissions or private data. A source evidence label is not our independent diagnosis.