What car years-to-avoid lists get wrong

Where the lists come from

Search any model with "years to avoid" and you'll find a list. Most of them lean on the same public source this site uses: owner complaints filed with NHTSA. It's a good source. It's also easy to misread, and a list that ranks years by total complaints misreads it in predictable ways.

Mistake 1: counting a model year's complaints without asking how old it is

A car collects complaints for as long as it's on the road. A model year that's been out for nine years has had nine years of owners filing. One that came out last year has had one.

The effect is large. For the model years from 2016 to 2020 on this site, 74% of all complaints arrived three or more calendar years after the model year. Most of what goes wrong with a car gets reported long after the first owner's honeymoon.

You can see it in the medians. Here's the typical complaint count per vehicle page, by model year:

Newer years sit lower for one main reason: they haven't had time. A list that says "2023 is the best year" off raw totals is often just saying "2023 is recent."

This site's model pages correct for it by dividing each year's complaints by the calendar years it's been on sale, and they don't call the two newest years at all. That's not perfect (problems don't arrive at a steady rate), but it's fairer than a raw total.

Mistake 2: ignoring how many were sold

A best-selling model draws more complaints than a niche one because there are more of them. NHTSA itself makes this adjustment when it screens complaints for possible defects: its recall booklet says staff measure the number of complaints and the severity of outcomes against the number of vehicles made and how long they've been on the road.

Public complaint files don't include sales figures, which is why comparing complaint counts across different models is shaky. Within one model it's safer, since production changes less from one year to the next than it does between models. That's why this site's years-to-avoid calls compare each year against the same model's other years, and never against a different car.

Mistake 3: treating every complaint as equal

A complaint about a frozen touchscreen and a complaint about a transmission that failed on the highway count the same in a total. They aren't the same to you.

Before you trust a "worst year," open it and read what the complaints are about. On this site, each vehicle page groups complaints by problem and shows how many involved a crash, a fire or an injury. Symptom sheets carry owners' own words. A year with a high count built from infotainment glitches can be a better buy than a lower-count year with a pattern of engine failures.

NHTSA draws its own line here. Its recall booklet lists defects it doesn't treat as safety-related, among them radios that don't work, cosmetic paint flaws and excessive oil consumption. Those can still be expensive problems to live with, and they'll never be recalled.

Mistake 4: forgetting the fix

Some of a bad year's worst problems were recalled and repaired for free. If the car you're considering has had the recall done, that problem may be behind it. Run the VIN at nhtsa.gov/recalls and ask for service records. Our VIN recall guide explains what the search shows.

The reverse also holds. A problem can be widespread without ever becoming a recall, which is why this site's recall pages list the systems owners complain about most that no recall covers.

Mistake 5: assuming complaint counts are a clean measurement

Owners file after something went wrong. Nobody files to say their car is fine. So complaint data only ever measures failure, and it can be pushed around by attention. This site's method notes that counts can jump after news coverage or an investigation.

None of that makes the data useless. It means a spike is a reason to look closer, and a quiet year isn't a guarantee.

How to use a years-to-avoid page well

1. Look at the rate, not the total. Use complaints per year on the road, and ignore calls on the two newest years.
2. Open the worst years. Read which problems drive the count, and at what mileage they tend to show up.
3. Check the mileage against the car in front of you. A failure that usually hits around 100,000 miles matters less on a car you'll sell at 60,000, and more on one that's already at 90,000.
4. Check recalls by VIN for the specific car.
5. Get an inspection from an independent mechanic who knows the model's weak spots. Our used-car research guide covers the full process.

What a good year looks like

A good year isn't one with no complaints. Every model draws some. It's a year where the complaint rate sits below the model's typical rate, the complaints that exist are spread across minor issues, and nothing clusters into a pattern of serious failures. When you find one, it's still worth running the VIN. Averages describe a model year, and you're buying one car.