A man crouches at the front wheel of his parked car on the driveway in morning light, checking behind the rim.

Brakes, tyres, lights: the checks that change the odds

Not a service manual. What a soft pedal or a worn tyre actually does to a crash, when each defect announces itself, and the check that takes two minutes before a long drive.

Updated September 15, 2026 Beginner

Defects do not announce themselves at random. Each one has a moment when it becomes noticeable, and the moments are spread across a journey in a fixed order: some are visible while the car is cold and parked, some in the first hundred metres, some only at speed, and one is never noticeable at all and has to be read off a date stamped into the rubber.

That order is what this piece follows. It is not a service manual, and none of it replaces a mechanic. It is the set of things a driver can detect without tools, and what each one does to the odds if it is left alone.

Cold and parked

The tyres are worth more of your attention than everything else on the car put together, and they are easiest to judge before they have been driven on.

Pressure first. The Highway Code’s instruction is to check weekly, before the journey, while the tyres are cold, because warm tyres give a misleading reading. The reason this matters is not fuel economy. NHTSA’s study of tyre factors in the phase before a crash found that among tyres underinflated by more than a quarter of the recommended pressure, about a tenth were on vehicles that experienced a tyre problem in the pre-crash phase, against a much smaller share of correctly inflated tyres — a threefold difference in how often the tyre was cited as a critical event. The study is honest that underinflation is not the only cause of tyre problems, and it carries a caveat worth keeping in view: tyre-pressure monitoring had only just become mandatory when the data was collected, so almost none of the surveyed vehicles had it.

Tread next, and measured rather than eyeballed. In the United Kingdom the legal floor is written into the Construction and Use Regulations as at least 1.6 mm across the central three-quarters of the tread around the whole circumference, and Spain’s traffic authority states the same figure while adding that it is unwise to let tyres get that far. The NHTSA study puts a number on why: around a quarter of tyres in the lowest tread band were on vehicles with a tyre problem before the crash, against roughly a twelfth of tyres with a little more tread left. Nothing in the legal minimum promises wet grip. European type approval sets a minimum wet-grip index for a tyre when it is new; it says nothing about the same tyre four winters later.

Then the lights, the glass and the ground under the car. The Highway Code requires lights, indicators, reflectors and number plates to be kept clean and clear and headlights to be properly adjusted so they do not dazzle, and windscreens and windows to be free from obstructions to vision. Spain’s DGT notes drily that a blown headlamp carries a fine. Headlights are also the one item on this list with a clean crash number attached, though it belongs to the lamp and not to the cleaning: the Insurance Institute for Highway Safety found good-rated headlight systems associated with a nineteen percent lower rate of night-time single-vehicle crashes than poor-rated ones, across tens of thousands of crashes. That is a fact about how a car was designed, which you learn when you buy it and cannot change afterwards. What you can change is whether the lenses are dirty and whether both of them work.

A puddle under a parked car is information. So is the level in the reservoirs, which the Highway Code says to check at least weekly, adding the sentence that explains why it is on the list at all: low brake fluid may result in brake failure and a crash.

The first hundred metres

The brakes report on themselves as soon as you use them, and the vocabulary is worth knowing because it maps onto what a workshop will find.

A pedal that goes further than it used to, or sinks slowly while you hold it at a stop, is a hydraulic symptom rather than a wear symptom. The MOT inspection manual classifies a leaking master cylinder and a damaged or excessively corroded brake pipe as serious defects for exactly this reason, and low fluid in the reservoir with nothing visibly leaking usually means the fluid went into the pads as they wore. Either way it is the same day’s job, not next month’s.

A car that pulls to one side under braking is telling you the two sides are not doing equal work — a sticking caliper, a contaminated pad, an uneven tyre. Grinding is past the warning stage: the MOT manual treats a pad worn down to its wear indicator as a major defect and a pad worn below 1.5 mm as a dangerous one, and the noise arrives somewhere in between. A squeal on a cold damp morning that disappears after a few stops is usually surface rust and not the same thing.

The steering and the wipers report in the same stretch. A wiper that smears rather than clears is a visibility defect with a fixed cost and a two-minute fix, and the DGT’s advice on wipers is simply to replace them as soon as they lose effectiveness rather than at an interval.

At speed, and at night

Some faults only appear under load. A vibration through the wheel at a particular speed, a shudder felt through the pedal in a hard stop, a car that wanders on a straight road — all of these are the first honest information you get about wheels, discs and geometry, and all of them are easy to file away as normal because they arrive gradually.

Night is its own inspection. Driving the same familiar road in the dark tells you whether your beam pattern still reaches where it used to, whether one lamp has gone yellow, and whether the windscreen that looked clean in daylight is scattering every oncoming headlight into a haze. Interior glass gets dirty slowly and nobody notices until they are driving into low sun or oncoming traffic.

The check nothing on the road will tell you

Rubber ages whether or not it is used, and the car cannot report this. The date of manufacture is moulded into the sidewall. Michelin’s own guidance is that tyres should be inspected annually by a professional after five years in service and replaced ten years after the date of manufacture regardless of remaining tread or apparent condition. Britain’s standards agency has turned the same concern into a rule for one category of vehicle, prohibiting tyres more than ten years old on the front steered axle of a minibus. A caravan or a second car with plenty of tread and a fifteen-year-old date code is the common case, and it looks perfect.

And the coolant question

Coolant is where a preparedness habit and a postponed repair look identical from outside the car.

A bottle in the boot is a reasonable thing to carry on a long drive in a car that is known to use a small amount between services. What it must not become is the reason a leak never gets fixed. A sealed system that needs topping up every few weeks is losing coolant somewhere, and the same weekly check that the Highway Code asks for is what distinguishes a slow, stable, known consumption from a loss that is getting worse. The honest test is whether you know the rate. If you are topping up without knowing how often you did it last time, the bottle is not preparation.

For a fleet, the same checks become a record

What a driver does by habit, a company has to be able to prove. In the United States the federal rules make the driver’s pre-trip duty explicit: be satisfied the vehicle is in safe operating condition, review the last inspection report where one is required, and sign it to acknowledge the review and the certification that repairs were made. The report itself has to cover a named set of parts — service and parking brakes, steering, lighting devices and reflectors, tyres, horn, wipers, mirrors, coupling devices, wheels and rims, emergency equipment — which is as good a spine for any company’s check sheet as anything a consultant will sell. In Britain, the walkaround guidance states the same principle in one sentence — the driver is responsible for making sure the vehicle is safe to drive — and adds the part that fleets get wrong, which is that defects are reported in writing to the person responsible for fixing them.

The failure mode in a fleet is not that checks are not done. It is that they are done and not recorded, or recorded and not routed to anyone who can authorise a repair. A signed check sheet with a defect on it and no closing entry is worse than no sheet at all, because it proves the company knew.

What the numbers do and do not support

The statistic that gets quoted against everything above is NHTSA’s national crash causation survey, which assigned the critical reason to the driver in ninety-four percent of crashes and to a vehicle component’s failure or degradation in about two percent. Anyone using it to argue that maintenance is marginal is arguing against the study’s own text. It states that the critical reason is not to be read as the cause of the crash, and it states that no vehicle received a detailed inspection — the vehicle-related reasons were inferred from looking at cars externally, so only visible items were ever cited, and the authors write that the resulting statistics may not be representative of internal problems that led to crashes. Within that two percent, tyres and wheels account for the largest share and brakes the next largest, with confidence intervals wide enough that the ordering is the finding and the figures are not.

Which is roughly what a driver would guess. The parts of a car that fail visibly are the parts a driver can check.

What we cannot tell you

We cannot tell you how much a given check reduces your own risk, because the studies that exist measure populations and not people, and the honest ones say so. We cannot tell you when your pads need replacing without measuring them. And we cannot tell you how a defect will be treated if it turns out to have contributed to a crash — whether knowing about it and driving anyway affects what you recover is a question of the negligence rule where you are, which is in the data below.

The two minutes before a long drive

Walk round the car. Look at each tyre from the front rather than from above, look for a shine or a ridge on the inner or outer shoulder, and put a gauge on them if they have not been checked this week. Look at the ground where the car has been standing. Switch the lights on and walk round again, including the indicators and the brake lights with someone behind the car or a reflection to watch. Clean the glass inside as well as outside, and fill the washer bottle. Then, in the first quiet stretch of road, brake once, deliberately, harder than you need to, and pay attention to what the car does.

None of that finds everything. It finds the things that would otherwise find you at the worst moment, which is a lower standard and a useful one.

Rules in your jurisdiction

Deadlines, fault rules and minimum coverage differ by state and country. Pick yours to see the rules that apply to this topic.

Select a jurisdiction to see its rules.

Frequently asked questions

If almost all crashes are caused by drivers, does maintenance matter at all?

The figure people quote for this is NHTSA's national crash causation survey, which assigned the critical reason to the driver in 94 percent of crashes and to a vehicle component in about 2 percent. The study says twice over that this is not what it looks like. It states that the critical reason is not the cause of the crash, and it states that no vehicle was actually inspected — the vehicle findings came from looking at cars from the outside, so only visible things like tyres and brakes were ever cited, and the authors write that the statistics may not be representative of internal vehicle problems. A number produced by not inspecting vehicles is weak evidence that vehicles are fine.

Is 1.6 mm of tread legal, or safe?

In the United Kingdom and across much of Europe it is the legal minimum for a car: the Construction and Use Regulations require at least 1.6 mm across the central three-quarters of the tread around the whole circumference, and Spain's DGT repeats the figure. Legal and safe are different questions. NHTSA's pre-crash tyre study found that around a quarter of tyres in the 0 to 2/32 inch band were on vehicles that had a tyre problem in the pre-crash phase, against a much smaller share of tyres with more tread left.

Should I carry coolant in the boot?

Carrying it is sensible for a car that is known to use a little between services and is being watched. Needing it regularly is not a supply problem, it is a leak, and the tank going down means the coolant is going somewhere. The Highway Code's instruction is to check fluid levels at least weekly, which is the habit that tells you which of the two situations you are in.