How to Read NHTSA Consumer Complaints (And What to Believe)

NHTSA's Vehicle Owner Defect Report database (VOQ) is one of the largest pools of unmoderated, owner-reported, vehicle-specific safety complaints anywhere. As of this year, it contains more than 2.1 million individual complaints filed directly by vehicle owners, going back to the 1990s. It is a research goldmine — and also a source of misinformation if you read it without context. Here is how to make sense of what you find.

What a complaint actually is

A VOQ is a free-form report filed by a vehicle owner to NHTSA describing a safety concern. The owner enters their VIN, the date of incident, and a description of what happened. NHTSA does not investigate the truthfulness of the complaint — they collect it and make it public. The manufacturer is notified and may respond, but the underlying complaint stays in the database whether the manufacturer agrees with it or not.

This is by design. The point of the VOQ system is to surface emerging defects from owners' actual experiences. NHTSA's investigators then look at patterns across complaints to decide whether to open an Office of Defects Investigation (ODI) inquiry, which can lead to a recall.

It also means that any individual VOQ might be wrong. Someone might misattribute a maintenance issue (worn brakes) to a defect (brake failure). Someone might exaggerate. Someone might be testing the system. NHTSA leaves all of those in the database.

The fields that matter

Each VOQ has a few fields worth understanding:

NHTSA number / ODI number. A unique identifier for the complaint, useful when referencing it elsewhere.
Component. The standardized part of the vehicle the complaint is about — Brakes, Engine, Power Train, Air Bags, Seat Belts, etc. NHTSA assigns this from a fixed taxonomy. If you're looking for a pattern, filtering by component is the most useful first cut.
Date of incident. The date the failure occurred, not the date the complaint was filed. Useful for timeline analysis.
Mileage at incident. Often blank, but when present, tells you when in the vehicle's life the failure occurred.
Crash, fire, injury, death flags. Boolean fields the owner sets when filing. These are owner-reported, not verified, but they're the cleanest way to filter for safety-relevant complaints.
Description. The free-text narrative. This is where you get the actual story.

How to read the description

The description is unmoderated and uncopyedited. Many are typed in caps lock. Many have grammatical errors. Many were clearly typed in the immediate aftermath of an accident and read as such. None of that affects their reliability — what affects reliability is whether the description is consistent with other complaints on similar vehicles.

What to look for:

Specific failure modes. "Transmission slipped" is vague. "Transmission slipped from second to neutral while merging onto highway" is specific. Specific descriptions are more credible because they're harder to make up and easier to verify.
Repair attempts. Many descriptions include the dealer's diagnosis and attempted repair. "Dealer replaced solenoid B; problem returned 2 months later" tells you both that the symptom was real and that the manufacturer's first-line fix didn't solve it.
Multiple owners reporting the same chain of events. If twenty owners describe transmission slip → dealer replaces solenoid → problem returns, that's a credible pattern. If three owners describe a transmission slip and seventeen describe other things, the slip might be coincidence.

How to interpret complaint counts

Raw complaint counts are misleading. A model that sold 1.2 million units will naturally have more complaints than a model that sold 200,000 units, even if the failure rate is identical. To compare across models, you need complaints per 100,000 vehicles sold, not raw totals.

NHTSA doesn't publish that ratio directly, but most aggregators (including this site) calculate it. The more useful metric is to compare a model against itself across years — a 2018 of a particular model with three times as many transmission complaints as the 2017 of the same model is suggestive of a year-specific defect, even if the absolute numbers are small.

Worth knowing: complaints lag failures. Most owners don't file a complaint until something has gone seriously wrong, and many never file at all. NHTSA's complaint volume for a given model year continues to grow for years after the model goes off sale, as the cars age and accumulate failures.

Investigations vs. complaints vs. recalls. A complaint is one owner's report. An investigation is NHTSA opening a formal inquiry, typically after enough complaints accumulate to suggest a defect. A recall is the manufacturer (sometimes under NHTSA pressure) issuing a remedy. Most complaints don't become investigations. Most investigations don't become recalls. But every recall starts as a pile of complaints.

What complaints don't tell you

Three things to keep in mind:

Reliability ≠ safety. The complaint database is for safety issues. A model with reliability problems — frequent breakdowns, expensive repairs — won't necessarily generate many VOQs unless the failures are also safety-relevant. For reliability data, look at Consumer Reports, owner surveys, and used-car forums; for safety data, the VOQ database is authoritative.
Defect ≠ pattern of cause. Two hundred complaints about a transmission don't necessarily mean the transmission is the cause; it might be a sensor, an electrical harness, or a software bug. NHTSA's investigation process exists to figure out the actual cause.
Absence of complaints isn't proof of safety. Some models are so new there hasn't been time for complaints to accumulate. Some failure modes only appear at high mileage and won't show up in early-life data. Look at how old the data is, not just how many complaints there are.

Reading complaints as a buyer

If you're using the complaint database to evaluate a used car, here's a workflow that works:

Filter to the specific year and model you're considering.
Look at the component breakdown. Which categories have the most complaints?
For the top one or two components, read 20–30 actual descriptions. You'll quickly see whether they describe the same failure mode or different ones.
Note any complaints with crash, fire, injury, or death flags. Read those carefully — they often describe the failure mode at its worst.
Cross-check the patterns you've found against the recall list for that year and model. Has the manufacturer addressed any of them through a recall? Through a TSB?

If a recurring pattern in complaints has been addressed by a recall and the recall is marked complete on the specific VIN you're considering, you can usually consider that pattern resolved. If it shows up in complaints but no recall has been issued, it's something to ask the seller about and to budget for.

If you're filing your own complaint

The complaint database is more useful when more people file. If you've experienced a safety-relevant defect, file a VOQ at nhtsa.gov/recalls. Be specific. Include the date, mileage, what happened, what the dealer said, and what (if anything) was repaired. Use clear language. Avoid speculation about the cause; describe what you observed. The more concrete the report, the more useful it is to other owners and to NHTSA's investigators.

About the author Written by the RecallCheck editorial team. We work directly with NHTSA, FDA, CPSC, and EPA data sources to build the searchable databases on this site, and we write these guides to help everyday readers make sense of what the data actually says.