Car Reliability by Make, Model & Year
CarWhere analyzed 429,585 NHTSA owner complaints across 178 models from 32 brands. Because complaint totals are not adjusted for sales volume, we do not rank brands against one another. Instead, each model year is compared with other years of the same model to flag unusually high complaint severity — the years that deserve extra scrutiny.
Analysis by CarWhere Research · Updated August 28, 2026 · NHTSA data as of July 2026 · Methodology · NHTSA source data · Download the dataset (CSV)
Key findings
- Of the 2,831 model years analyzed, 609 rate “avoid” and 214 rate “best” — most model years are neither standouts nor problem years.
- 36 of 178 models carry no model year flagged “avoid” — though several of those simply have records too thin to rate, so a missing flag is not proof of a clean history.
- Complaint totals track volume, not durability: the highest totals belong to high-volume brands (Honda, Jeep, Toyota), the lowest to low-volume ones (Jaguar, Porsche, Genesis) — the reason a raw count can't rank brands.
- Reliability is a model-year question, not a badge: 142 of 178 models carry at least one flagged year, and nearly every brand has both flagged years and clean years within its lineup.
How the 2,831 model years break down
Every model year gets a verdict computed against that model's own history — never against other brands.
| Verdict | Model years | What it means |
|---|---|---|
| Best | 214 | The model's lowest severity-weighted complaint-rate years with enough history to be confident — the two newest years are never rated best. |
| Solid | 557 | Complaint rate in line with or below that model's own median year. |
| Caution | 386 | Elevated against the model's own median, but below the avoid threshold. |
| Avoid | 609 | Severity-weighted complaint rate among the highest in that model's own history (≈1.6× its median or worse, minimum 40 lifetime complaints). |
| Insufficient data | 1,065 | Too few complaints or too little time on the road to rate fairly — never counted as evidence a year is clean. |
Largest complaint spikes vs. a model's own baseline
The models whose worst year logged the highest severity-weighted complaint rate relative to that model's own median year — the only cross-model comparison this method supports, because each multiplier is normalized within its own model before comparing.
| Model | Worst year vs. its own median | Years flagged “avoid” |
|---|---|---|
| Toyota Prius | 18.8× | 2008, 2009, 2010, 2011, 2012, 2013, 2014 |
| Ram 3500 | 9.4× | 2014, 2017, 2018, 2019, 2020, 2022 |
| Hyundai Sonata | 9.3× | 2011, 2012, 2013, 2014, 2015, 2016, 2017 |
| Jeep Compass | 9× | 2014, 2016, 2017, 2018, 2019, 2020, 2022 |
| Buick Regal | 8.7× | 2011, 2012, 2013 |
Where problems concentrate
The vehicle systems drawing the most federal service-bulletin (TSB) filings across all 178 models — a volume signal of where manufacturers issue the most fix guidance, counted from NHTSA component categories.
How to read this. These are NHTSA complaint and recall patterns by model year — not a sales-adjusted reliability ranking. The 429,585 figure is owner complaints (recalls are counted separately). To compare brands head-to-head you would need to normalize for vehicles sold, model age, and miles driven, which no public dataset provides by model year.
Every brand in the dataset
Alphabetical — not a ranking. Complaint counts are raw federal filings, not adjusted for sales volume.
| Make | Models analyzed | Owner complaints | Models with a flagged year |
|---|---|---|---|
| Acura | 5 | 4,779 | 4 of 5 |
| Audi | 6 | 3,106 | 6 of 6 |
| BMW | 6 | 4,992 | 5 of 6 |
| Buick | 6 | 5,823 | 5 of 6 |
| Cadillac | 6 | 6,267 | 6 of 6 |
| Chevrolet | 6 | 28,015 | 6 of 6 |
| Chrysler | 6 | 14,380 | 4 of 6 |
| Dodge | 6 | 23,257 | 6 of 6 |
| Ford | 2 | 34,981 | 2 of 2 |
| Genesis | 6 | 813 | 3 of 6 |
| GMC | 6 | 17,437 | 5 of 6 |
| Honda | 6 | 44,822 | 6 of 6 |
| Hyundai | 6 | 31,939 | 6 of 6 |
| Infiniti | 6 | 1,777 | 3 of 6 |
| Jaguar | 5 | 619 | 2 of 5 |
| Jeep | 6 | 39,629 | 6 of 6 |
| Kia | 6 | 26,115 | 6 of 6 |
| Land Rover | 6 | 3,539 | 5 of 6 |
| Lexus | 6 | 2,583 | 2 of 6 |
| Lincoln | 5 | 2,072 | 5 of 5 |
| Mazda | 6 | 7,478 | 4 of 6 |
| Mercedes-Benz | 6 | 9,323 | 6 of 6 |
| Mini | 6 | 1,447 | 2 of 6 |
| Mitsubishi | 6 | 2,965 | 4 of 6 |
| Nissan | 6 | 25,998 | 6 of 6 |
| Porsche | 5 | 660 | 1 of 5 |
| Ram | 5 | 15,323 | 3 of 5 |
| Subaru | 4 | 8,511 | 4 of 4 |
| Tesla | 6 | 12,393 | 4 of 6 |
| Toyota | 6 | 33,748 | 6 of 6 |
| Volkswagen | 6 | 13,151 | 6 of 6 |
| Volvo | 3 | 1,643 | 3 of 3 |
Reliability by brand
How we built this
Sources: the public NHTSA owner-complaints API and recalls API, with make/model names normalized via NHTSA vPIC. For each model year we severity-weight complaints (crash/fire/injury and safety-critical systems count double), annualize by years on the road, compare against the model’s own median, and fold in recalls — mapping to an avoid / caution / solid / best verdict. Retrieved July 2026. Model years 2008–2025.Full methodology and exact scoring →
Limitation: complaint counts are raw federal filings, not adjusted for how many of each model year were sold or how far they’re driven — so they compare years within a model, never brands against each other.
Suggested citation
CarWhere Research. "Car Reliability by Make, Model & Year: NHTSA Complaint and Recall Analysis." Based on 429,585 NHTSA owner complaints covering model years 2008–2025. Updated August 28, 2026. https://www.carwhere.com/research/reliability
This dataset is free to reuse under CC BY 4.0 with attribution to CarWhere Research and a link to this page.
Download the dataset (CSV) — per-model complaint totals, flagged years and verdict counts for every brand we cover (version 2026-08-28).
Frequently asked
What is the most reliable car brand?
This dataset can't answer that. It ranks NHTSA owner-complaint and recall patterns, and raw complaint totals track how many vehicles a brand sold and how long they've been on the road — not durability. The highest totals here belong to high-volume brands (Honda, Jeep, Toyota); the lowest to low-volume ones (Jaguar, Porsche, Genesis). We compare each model year against other years of the same model, and deliberately do not rank brands against each other.
Can NHTSA complaints measure reliability?
Partly. NHTSA owner complaints are a public federal record of problems owners reported, weighted here toward safety-critical failures. They are a strong signal of which model years drew unusually high owner concern, but they are not a sales-adjusted dependability survey like Consumer Reports or J.D. Power. Use them alongside those.
What makes a model year get flagged?
A model year is flagged "avoid" when its severity-weighted complaint rate is among the highest within that model's own history — roughly 1.6× the model's median or worse, with at least 40 lifetime complaints so a low-volume year can't be flagged on a handful of reports. Full formula on the methodology page.
How does CarWhere flag years to avoid?
For each model year we take NHTSA owner complaints, weight severe ones (crash, fire, injury, safety-critical systems) double, divide by years on the road, and compare against the model’s own median, then fold in recalls. The result maps to an avoid / caution / solid / best verdict. The method is a fixed, published algorithm.
How should I check a specific used car?
Start with the model’s reliability page here to see which years to favor, then run that exact VIN through a report for its open recalls, complaint history, and what verified buyers paid before you negotiate.
Ready to find the right car?
Use the reliability data to choose the right model year—then compare live inventory and see what buyers actually paid.
See What Buyers Paid
Real out-the-door prices from verified buyers — know a fair deal before you negotiate.
Keep Researching
Years to avoid, dealer fees, incentives and more from CarWhere Research.