Used Car Reviews

Not how it drives on a launch event. What breaks, at what mileage, and what the fix costs. Each guide covers a model and a range of years, because the difference between a good one and a bad one is usually a single production change.

How these used car guides are written

Model-by-model buying guides built on what breaks, not on how a car drives. The questions here are the ones asked with money on the table: which years to avoid, what the known fault costs, what to check on a viewing, and whether a high-mileage example is worth the risk.

Where the data comes from

NHTSA's public complaints database, queried directly and counted here rather than quoted from someone else's summary. Every guide states the filter it applied and the date it was taken.

Recalls, warranty extensions and service campaigns are linked to the manufacturer's own filing.

The metric, and its flaw

Raw complaint counts rank sales volume, not reliability. So the measure used is the share of a model's own complaints that describe an expensive failure. Volume cancels out.

It has a known flaw: one huge non-drivetrain recall flatters a model's score. Where that happens, the guide names it.

What we don't do

We haven't stripped these cars. Where a guide describes a failure mode, that's published engineering; where it describes a maker's response, that's a document you can open.

We don't calculate complaints per vehicle sold. Nobody publishes the sales figures free, and inventing a denominator would be worse than leaving the gap visible.

Found a figure here that doesn't match what you were quoted, or think something on this page is wrong? Tell us. Corrections get read first and the page gets dated when it changes. If you want to know how these guides are put together before you trust one, that's written down too.