Vehicle Components Most Linked to Crashes, Fires, and Fatalities

Why some complaints become crashes

Most of the records in the NHTSA Office of Defects Investigation database describe an annoyance, a warning light, or a part that wore out early. A small fraction describe something far worse: the owner was driving when the part failed, and the failure put the vehicle, its occupants, or other road users in danger. The difference between those two kinds of report is what this analysis isolates. Rather than rank components by how many complaints mention them, we rank them by how often a complaint about that component also records a crash or a fire in the same incident.

That distinction matters because raw complaint volume is dominated by parts people touch constantly. Across the 15 component systems shown here, owners filed 1,161,587 complaints, and those same records carry 94,167 crash flags and 30,075 fire flags. A high complaint count alone tells you a subsystem is common or fragile. A high crash count tells you the subsystem fails in a way that takes the driver by surprise at speed.

Which car parts cause the most crashes?

By absolute crash count, the subsystem at the top of this dataset is AIR BAGS, which appears in 23,393 complaint records that also report a crash. That is a notable result on its own: air-bag complaints reach NHTSA both when a bag deploys without cause and when one fails to deploy in a collision, so a crash flag attaches to a large share of them. The next tier is dominated by the systems a driver relies on to keep control, speed control, braking, and steering, where a failure mid-journey is almost definitionally a crash risk rather than a cosmetic defect.

The chart below renders the crash-linked count for each subsystem directly from the query result. Bars are sorted high to low, so the systems that most often escalate into a collision sit at the top.

The components that escalate most

Absolute counts reward the subsystems people complain about most. A fairer question is conditional: given that someone filed a complaint about a part, how likely is that same record to carry a crash flag? We call that the escalation rate, crashes divided by complaints for each component. It separates the parts that fail loudly but harmlessly from the parts that fail dangerously.

On that measure the leader is AIR BAGS:FRONTAL, where 14,871 of 20,105 complaints, about 74.0% - also record a crash. Contrast that with a system like the electrical system, which generates one of the largest complaint piles in the whole database yet escalates to a crash in only a small share of cases, because most electrical faults strand a vehicle rather than wreck it. The table below shows both views side by side, so a subsystem near the top of the crash column but low on the escalation column reveals itself as high-volume-but-survivable, while the reverse pattern marks a subsystem that is comparatively rare but unforgiving.

Fire risk versus crash risk

Crashes and fires are different failure modes, and they do not track the same components. A crash usually follows a sudden loss of control, a steering or braking input that does not produce the expected response. A fire usually follows a thermal or electrical fault that has nothing to do with how the car is being driven. Ranking the same subsystems by fire flags instead of crash flags reorders the list almost completely.

By fire count the leader is ELECTRICAL SYSTEM, with 9,814 complaints that also report a fire. Electrical and engine-related systems cluster at the top of the fire ranking precisely because they manage current and heat, the two ingredients of an under-hood fire, whereas they sit lower on the crash ranking. The second chart makes the contrast visible: a reader comparing the two charts will see air bags and speed control dominate one and electrical and engine systems dominate the other, with very little overlap at the very top.

Where injuries and fatalities concentrate

The harshest column in the table is the death count, and it does not simply mirror the crash column. Among the systems shown here, the most fatalities attach to AIR BAGS, with 799 complaint records that also report a death. Across the visible set, 3,396 death flags are spread unevenly: a subsystem can produce many minor crashes without producing many fatalities, while another produces fewer crashes but a higher fatal share because the failures it causes happen at highway speed or disable a primary safety device. Speed-control and air-bag systems carry disproportionate fatality weight for exactly that reason, one removes the driver's ability to slow down, the other removes the occupant's last line of protection in the collision that follows.

How to read these numbers

A few cautions keep these figures honest. First, the crash, fire, injury, and death flags are self-reported fields on a complaint, not adjudicated findings. An owner who marks "crash" is telling NHTSA that a collision occurred in connection with the complaint; the agency does not confirm causation, and a single incident may be reported by several parties. Second, the counts are absolute, not normalized by how many vehicles carry each subsystem or how many miles they travel. Every modern car has brakes, steering, and an electrical system, so those subsystems generate large totals partly by being universal. Third, the same physical defect can be filed under several component labels depending on how the owner or the intake clerk categorized it, which is why closely related rows, frontal air bags versus air bags overall, hydraulic brakes versus brakes overall, appear separately. Read the table as a map of where danger concentrates, not as a precise hazard rate for any one model.

The right way to use a page like this is comparative. If you are researching a specific vehicle, follow the component links to see which makes and models drive each subsystem's totals, then weigh a high crash or fire share more heavily than a high raw complaint count. A part that fails often but safely is an inconvenience; a part that fails rarely but dangerously is the one worth checking the recall history for.

Methodology and data notes

The figures on this page cover every NHTSA component category with at least 500 complaints, so the ranking is not skewed by rare categories where a handful of severe reports would produce a misleadingly high escalation rate, ranked by the sum of crash and fire counts. The escalation rate shown in the table (crash count divided by complaint count) is always computed from the same two numbers shown beside it, so it can never drift out of sync.

The numbers on this page reflect the current NHTSA dataset as of your visit and update automatically as new data is published; nothing is hardcoded. The component links carry through to per-subsystem detail pages, and the methodology page documents the source file vintage and refresh cadence.