Most of the vehicle safety record counts events. Crashes, injuries, deaths, fires — things with a measurable aftermath.
54,604 complaints describe none of that. They report something that almost happened, and the owner filed anyway. That is roughly one complaint in twenty.
| Component | Near-miss complaints |
|---|---|
| Power train | 5,694 |
| Steering | 4,551 |
| Engine | 3,373 |
| Service brakes | 2,698 |
The ordering is different from the one crashes produce, and the difference is informative.
A near miss is a failure with a witness
These reports share a structure. The car did something unexpected, the driver had to react, and the reaction worked. Nothing was damaged, so nothing was recorded anywhere except in the complaint.
That makes them the only part of the safety record describing failures that did not cost anything. They are also the part most likely to be under-reported, because there is no insurance claim forcing anyone to write it down.
The power train tops the list because its failures happen in traffic and require immediate judgement: a transmission that drops out of gear while overtaking, a sudden loss of drive on a slip road. The driver gets one decision, and in these 5,694 cases it went the right way.
Steering produces fewer crashes than near misses
Steering is second here with 4,551, well above brakes.
That inverts the usual ranking. In crash-involved complaints, brakes rank higher than steering. The reason is what each failure gives the driver to work with.
A brake failure removes an option and gives nothing back — there is no technique that replaces stopping distance. A steering fault usually changes effort rather than removing control: assist that cuts out mid-corner leaves the wheel heavy but still connected, and a driver who holds on keeps the car. Hence more near misses and proportionally fewer crashes.
Why it matters that these exist
Regulators open investigations on patterns, and a pattern needs volume. A defect that produces one crash and fifty near misses looks small in crash data and obvious in complaint data.
This is the argument for filing one. An owner whose car lurched in traffic and who managed it has nothing to claim and no reason to fill in a form, and that is exactly the report that turns a coincidence into a trend. Complaint counts multiply as a model ages, and the early ones — the ones filed before anything was damaged — are the ones that arrive in time to matter.
Reading this as a buyer
When you check a model's complaint history, the near misses are the part worth reading rather than counting. A summary that says a car lost drive on a motorway is a more useful description of a defect than a crash report, because it says what the fault actually does.
The counts tell you how common something is. The text tells you what it will feel like. Both sit in the same record, and most people only ever look at the first.
Frequently asked questions
What counts as a near-miss complaint?
A report where the car failed, the driver reacted, and nothing was damaged. Across the record, 54,604 complaints contain language describing something that almost happened — roughly one in twenty.
Why does steering produce more near misses than brakes?
Because the failures differ in what they leave the driver. A brake failure removes stopping ability with no substitute. A steering fault usually makes the wheel heavy rather than disconnecting it, so a driver who holds on often keeps control — producing a near miss instead of a crash.
Is it worth filing a complaint when nothing was damaged?
Yes, and these are arguably the most valuable reports in the system. Investigations are opened on patterns, and a defect that causes one crash and fifty near misses is invisible in crash statistics but clear in complaint data.
Denis Kataev is a serial solopreneur and the founder of ForCar. With 15+ years in software engineering and 10 years in SEO, he builds data-driven products end to end — backed by a sharp eye for design. At ForCar he mines proprietary vehicle datasets, turning raw numbers into buying advice you can actually trust.
Keep reading
All articles →
The Passenger Side Was Tested Fifteen Years After the Driver's. It Scores Better.
For over a decade, one side of the car was crash tested and the other was assumed to match. When the assumption was finally checked, the results said something unexpected about how safety engineering responds to being measured.
88% of Cars Failed the Whiplash Test in 2000. Since 2020, Not One Has.
Most safety measures improve slowly and never finish. This one finished. Tracing head restraint ratings across two decades shows what it looks like when an industry solves something completely.
11,447 People Were Injured by a Car That Never Crashed. 1,492 of Those Reports Name the Airbag.
Injury data is usually read as crash data. A quarter of the injuries in the complaint record happened with no impact of any kind, and the components responsible are not the ones you would expect.
Mark ForCar as a preferred source and our data shows up first in your Google results. One click, no account.