Timing Belt Replacement: The One That Ends the Car

There is one item on a service schedule that does not behave like the others. Skip an oil change and the engine wears faster. Skip a coolant change and the corrosion inhibitors run down. Skip brake fluid and the pedal goes long on a hot descent. All of those are gradients: the neglect accumulates, the bill grows, and there is usually a point at which somebody notices and it gets dealt with, more expensively than it needed to be but not catastrophically.

A timing belt is not a gradient. It is a cliff. It runs perfectly right up until the moment it does not, and on the majority of engines built in the last few decades the moment it stops is the moment the engine destroys itself. That single property — total, immediate, unwarned — is why a used-car site has to treat this differently from every other line on the maintenance list, and why the useful version of this page is less about the repair than about how you find out whether it has been done.

What the belt is actually holding together

Start with what is being kept in step, because the reason this failure is so total lives in the mechanics rather than in the belt.

A four-stroke engine completes its cycle — intake, compression, power, exhaust — over two full turns of the crankshaft. Each valve opens once in that cycle. So the camshaft that operates the valves has to turn at exactly half crankshaft speed, and not approximately half: exactly, continuously, for every revolution the engine ever makes. More than that, it has to hold a fixed angular relationship to the crank. The intake valve opens at a particular point in the piston’s travel and closes at another. Move that relationship by a few degrees and the engine runs badly. Move it further and it will not run at all.

Something has to enforce that relationship. On an overhead-cam engine the camshaft sits at the top of the cylinder head and the crankshaft sits at the bottom of the block, a long way apart, and the thing spanning that distance is either a toothed rubber belt or a metal chain.

Position, not power

This is what makes a timing belt a different class of component from the ribbed belt on the engine’s outer face, and it separates them more usefully than simply noting that they live in different places.

An accessory belt works by friction. It is pressed against smooth pulleys hard enough to drag them round, and a small amount of slip is designed into that arrangement — it is why a tensioner exists, and it is why a belt under a heavy load can squeal without anything being broken. Slip on that drive is a loss of efficiency. It is a nuisance.

A timing belt works by mesh. Its inner face carries teeth that sit in matching teeth on the crankshaft and camshaft sprockets, and it is not permitted to slip at all. Nothing about the job is negotiable by a few per cent. Jump one tooth and the cam phasing shifts by a fixed, substantial amount — the engine loses power, runs roughly, may set a fault for the camshaft and crankshaft sensors disagreeing with each other, and on some engines will not start. Jump several and the valves are being opened at times the pistons have not been told about.

The accessory drive is a genuinely important system with its own failure modes and its own inspection routine, and it is covered properly in our guide to what the serpentine belt drives and how to check one, which owns that subject on this site. The short version of why they get confused: they are both belts, both made of rubber, both at the front of the engine, and both get described by owners as “the belt”. The short version of why the confusion matters: one of them is a condition item you can look at in a car park, and the other is a scheduled item you cannot see at all.

Why a belt rather than a chain in the first place

It is worth knowing why manufacturers ever chose rubber for a job this consequential, because the answer explains the interval.

When camshafts moved from the block up into the cylinder head, the drive had to reach much further. A toothed belt does that well. It is light, so it puts less inertial load on the drive. It is quiet, and cam drive noise is one of the harder noises to engineer out of a car. It needs no lubrication, so it can run outside the engine’s oil in a dry compartment behind a cover, which simplifies sealing. It is cheap to make and cheap to fit on the production line. And a belt tolerates a long unsupported span better than a chain does.

Every one of those is real. The cost of them is that the belt is a consumable. Rubber and its reinforcing cords fatigue, harden with heat, and age chemically whether the car is driven or parked. A chain does not have that problem in the same form, which is why the industry has drifted back towards chains — and, as the next section argues, why that drift has been oversold to buyers as a solved problem.

One modern variation deserves naming, because it is now common on used cars and it surprises people. A number of recent engines use a belt that runs inside the engine, immersed in the oil, in place of a chain. The reasoning is sound enough: an oil-bathed belt is quieter than a chain and reduces friction. The consequences for an owner are not trivial. The belt’s condition becomes tied to the condition and specification of the engine oil, so an oil schedule that would merely be sloppy on a conventional engine becomes a timing-drive question. When these belts degrade they can shed material into the oil, and that material has somewhere to go. If you are looking at a car with this arrangement, the oil history is not a general indicator of care — it is direct evidence about the timing drive, which is why our guide to reading a car’s oil-change history is worth more on those engines than on almost any other.

Interference: the fact that changes what a failure costs

Everything above is true of any engine with a belt. What decides whether a broken belt is an inconvenience or the end of the vehicle is a single design characteristic, and it is not something you can see, hear, or infer from the badge.

Inside a cylinder, the piston sweeps up towards the head and the valves drop down into the chamber. Whether those two paths overlap is a choice the designer made. In a non-interference engine, the piston at its highest point still clears the valves at their fullest opening, with room to spare, under any phasing. Remove the timing drive on such an engine and the camshaft stops while the crankshaft keeps turning, and nothing touches anything. The car coasts to a halt, gets recovered, has a belt fitted, and drives away.

In an interference engine, those paths overlap. Under correct timing they are never in the same place at the same moment — the valve has lifted and returned before the piston arrives, over and over, thousands of times a minute, with a clearance measured in fractions of an inch. The belt is what guarantees the choreography. Take it away and the choreography ends. The valves stop where they are; the pistons continue at whatever speed the car was doing; and they meet.

Why most engines are the dangerous kind

The instinct is to hope you have the safe sort. The odds are against it, and the reason is efficiency rather than carelessness.

Clearance costs performance. Guaranteeing that a piston can never reach a valve means either lowering the compression ratio, which costs efficiency and power, or machining relief pockets into the piston crowns, which disrupts the combustion chamber shape and costs some of the same things. Meanwhile every pressure on engine design for decades — emissions, fuel economy, power from smaller capacities, turbocharging, variable valve timing — has pushed compression ratios up and valve timing towards more aggressive overlap. Those pressures point one way. The result is that interference layouts are now the ordinary case rather than the exception, and treating a modern engine as probably safe is the wrong default.

There is a further trap in how people reason about this. Interference is a property of the engine, not the model. A single model line can be sold with several engines across its production run, some belt-driven and some chain-driven, some interference and some not, and the body sitting in front of you tells you nothing about which one is under the bonnet. This is the reason the check has to be done against a specific vehicle rather than a specific car name.

How a buyer establishes it, in order of reliability

• The owner’s handbook and the manufacturer’s schedule for that engine. If a timing belt replacement interval is listed, the engine has a belt. If no such interval exists anywhere in the schedule, it almost certainly has a chain. This is the first question, because it decides whether the rest matters.
• A franchised dealer’s parts department, against the vehicle identification number. A parts system resolves the VIN to the exact engine and will tell you in a minute whether a timing belt kit is listed for it. This costs a phone call and is the single most reliable answer available to a private buyer.
• The engine code. Once you have the engine code rather than the marketing name, technical references and a specialist independent workshop for that make can usually tell you whether the layout is interference. Workshops that see one make all day know these engines individually.
• Owner communities for the specific engine. Useful and free, and worth reading for the pattern rather than for any single post. Treat it as corroboration, not as the answer.

If you can only establish one of those two facts, establish the belt-or-chain question first. An engine with a belt and an unknown interference status should be treated as interference until somebody proves otherwise, because the cost of being wrong in that direction is the entire engine and the cost of being wrong in the other direction is one unnecessary service.

The honest position on timing chains

Sellers and listings use the phrase “timing chain, so no belt to worry about” as though it settles something. It settles one thing and opens another, and a buyer is better served by the whole picture.

What is true: a chain is not a scheduled replacement item on most engines. It carries no interval, it does not age chemically the way rubber does, and it does not usually snap. The specific failure this page is about — a sudden, total loss of the timing relationship at speed — is genuinely rarer on a chain-driven engine. That is a real advantage and it should be counted as one.

What is also true is that chains wear, and the way they wear is unfamiliar to people who expect a metal part to be permanent. A chain is a long assembly of pins, bushings and plates, and every one of those joints wears microscopically. Multiply a tiny amount of wear at each joint by the number of joints and the chain gets longer. It is usually described as stretching, which is a slightly misleading word — the metal is not elongating, the joints are wearing — but the effect is what it says.

A hydraulic tensioner takes up that slack automatically, and it does so successfully for a long time. Then it reaches the end of its travel. Alongside it, the plastic guides that keep the chain running where it should wear grooves into themselves, and old plastic eventually cracks and breaks up. The pieces go somewhere.

The symptoms a buyer can actually use are these. A rattle from the timing end of the engine during the first seconds of a cold start, before oil pressure has built up enough to pump the tensioner out, is the classic tell, and it is why finding the car unstarted matters on a chain-driven engine as much as on a belt-driven one. Progressive loss of performance and economy, as a lengthened chain retards the cam timing a little at a time. And a stored fault for the camshaft and crankshaft position sensors disagreeing about where the engine is, which is the electronic version of the same finding.

Then the part nobody mentions in the listing. A chain job, when one is needed, is usually the more expensive of the two. A belt lives behind a plastic cover on the outside of the engine, and reaching it is a matter of removing things that are in the way. A chain lives inside the engine, behind a cover sealed to the block and often to the sump, running in oil. Getting to it can mean the sump off, the accessory drive off, sometimes the engine mount unloaded and on some layouts the engine out of the car. The parts are a chain, guides, tensioner and sprockets, and the sealing job on the way back together is not trivial.

There is also a maintenance link that a belt does not have. A chain is lubricated by engine oil, and it wears faster on oil that is old, wrong or persistently low. On an engine with a reputation for chain wear, the oil history is not a general character reference for the previous owner. It is evidence about the specific component.

What actually goes into the job

A timing belt replacement is not a belt replacement, and understanding why is the difference between reading an invoice and merely holding one.

Almost all of the cost of this job is access. Reaching the timing covers on a transverse engine commonly means the wheel off, the arch liner out, the accessory belt off, the crankshaft pulley off, and on many cars the engine supported from above or below while a mount is unbolted and moved out of the way. Once that is done, everything behind the covers is right there. Once the covers go back on, everything behind them costs the whole access job again to reach.

That single fact dictates what a competent shop fits while it is in there.

• The belt itself, always new. A timing belt is never reused once its tension has been released, and a shop that offers to refit an old one is telling you something about the shop.
• The tensioner. The component that sets and holds the belt’s tension, and on many engines an automatic unit with a bearing and a spring or hydraulic element in it. It wears on the same clock as the belt, and a seized tensioner bearing is one of the ways belts are destroyed well before their interval.
• The idler pulleys. Bearings whose only job is to guide the belt around obstacles and hold it against the sprockets. Same clock, same argument, and a seizure has the same consequence.
• The crankshaft and camshaft oil seals, on many jobs. These are not being replaced for the sake of it. A weeping cam or crank seal drips oil onto a belt that is designed to run dry, and oil-contaminated rubber swells, softens and lets go early. A new belt run past an old leaking seal has had its life shortened before the car leaves the ramp.
• Any single-use fasteners the manufacturer specifies — stretch bolts on the crank pulley or the tensioner on some engines, which are torqued to yield and are not reusable.
• The water pump, wherever the pump takes its drive from the timing belt. This is the one that gets read as an upsell and almost never is.

Why the water pump goes in with it

Keep the logic narrow, because it is simple and it is entirely about geography. On a large share of belt-driven engines, the water pump takes its drive from the timing belt rather than from the accessory belt at the front. That means the pump sits behind the same covers, is reached by the same dismantling, and is exposed the moment the belt is off.

The pump is a wear item with a bearing and a seal in it. If it is left in place and gives up eighteen months later, the entire access job is repeated, and the belt that was fitted at the same time is scrapped in the process because it has to come off again and it cannot be reused. Two jobs, two lots of labour, one of them entirely avoidable. That is the whole argument, and it holds regardless of what the pump costs.

There is a second-order reason as well. When a belt-driven pump seizes, it does not fail politely. A seized pulley behind the covers is exactly the sort of event that destroys the belt immediately, which on an interference engine converts a cooling fault into an engine failure. The relationship between the cooling system and the timing drive on these engines is closer than most owners realise, and our guide to what a rising temperature gauge is telling you covers the cooling side of that relationship in its own right.

The corollary matters for reading paperwork: on those engines a genuine timing belt invoice usually has a coolant line on it, because the system had to be drained and refilled. An invoice claiming a belt job on a pump-driven-by-belt engine with no coolant anywhere on it is worth a question.

The interval, and the half of it that gets ignored

One published service table runs across this site, and this component occupies its last line as the longest-interval item on it. The figure given there, for engines that have a belt at all, opens at 60,000 miles and closes at 105,000, with a note beside it saying the work is not optional where the engine is an interference design. Every line of that table, and the argument behind each one, sits on our page about the maintenance a used car genuinely needs.

Two things about that range are worth drawing out.

The first is how wide it is, and why the width is not vagueness. Manufacturers set this figure themselves, per engine, and they vary enormously in how conservative they are. The figure that governs the car you are actually looking at is the one in its own schedule, and no other version of it counts. A published range tells you which band the answer falls in. It does not tell you the answer.

The second is that the mileage is only half of the specification. Every manufacturer that publishes a belt interval publishes a time limit alongside it, expressed in years, and the instruction is always whichever comes first. This is not a formality. Rubber ages by oxidation and by heat cycling, and a belt in a car that sits on a driveway is ageing at close to the rate of a belt in a car that is being driven. The engine bay still gets hot when the car runs, and cools when it stops, over and over, for years.

It is also worth seeing where this item sits relative to the rest of the schedule, because it explains why competent shops bundle work at a particular point in a car’s life. On the same table, the spark plug window opens at 60,000 miles and shuts at 100,000, while coolant carries a limit of five years or that same 100,000, whichever lands first. Both of those overlap heavily with the belt window. A car arriving for a belt at the top of its range is a car that is plausibly due for plugs and certainly due for coolant, and a workshop that raises all three at once is not padding the bill. It is reading the same table you are.

Why you get almost no notice

Most components on a car degrade in ways you can perceive. Brakes get less effective and start to make noise. Tyres lose tread you can measure. A failing bearing announces itself for months. Even a slipping accessory belt squeals at you.

A timing belt does very little of that, and being straight about it is more useful than manufacturing a symptom list. The belt is enclosed, so it makes little noise you can attribute to it. It is not inspectable in any meaningful way on most engines — and where a cover can be lifted to expose part of it, what you see is the outer face of a section of belt, which tells you almost nothing about the state of the cords inside it or the roots of the teeth. A belt approaching the end of its life looks broadly like a belt at the start of one.

What you can sometimes catch is not the belt failing. It is something around the belt failing, which then takes the belt with it.

• A whine, growl or rattle from the timing cover area that follows engine speed and ignores what you switch on. That is a bearing — a tensioner or an idler — and it is the most useful audible warning available in this whole subject, because a seizing pulley is one of the commonest ways a belt is destroyed before its time.
• Oil or coolant in the vicinity of the covers. A leaking cam or crank seal, or a weeping belt-driven pump, is putting fluid onto a belt that must run dry. Contaminated rubber does not last.
• Rough running, a hard start, or a fault for the camshaft and crankshaft sensors disagreeing. This is the aftermath of a belt that has already jumped a tooth, which happens on a worn belt over worn sprockets or a belt that has lost tension. It is a warning in the sense that the failure has begun and has not yet completed.
• Fine black rubber debris found near the lower cover or in the vicinity of the crank pulley, where a belt has been shedding material against something it should not be touching.

Read that list honestly and the conclusion is uncomfortable: none of it is reliable, and the most common experience of a timing belt failure is no experience at all until the engine stops. That is precisely why the item is on a schedule. A wear component that cannot be assessed on condition has to be replaced on time, and the interval exists to substitute for an inspection that is not available.

What sets the bill, and why no figure appears here

We do not print repair prices, and the reasons are stated in full elsewhere on the site. There is a reason specific to this job that is worth adding, because it is the strongest version of the argument anywhere on this site.

Two numbers are involved in a timing belt decision, not one. The first is what the job costs, which varies more between engines than almost any other scheduled service — the same nominal task is a straightforward morning on one car and the better part of a day of dismantling on another, and a single range would hide exactly that. The second number is what the failure costs, and that one has no meaningful range at all: on an interference engine it is bounded above by what the vehicle is worth, because past a certain point nobody repairs it. Publishing a figure for the first while the second is sitting there unquantified would tell a reader the least useful half of the story.

What is worth setting out is the list of things that move the first number, because every one of them can be pinned to a particular car ahead of any quotation.

Look down the last column. Every one of those is answerable by a phone call before you commit to anything, and the answers turn an unspecified worry into a number you can put on the table.

What the federal record can and cannot say about this

Two archived federal datasets sit behind this site, and it is worth being direct about how they relate to a timing belt: neither of them can see one. Neither absence is a hole in the record. Each follows from what that dataset was built to collect, and working out why is genuinely useful to a buyer.

The complaint archive, and a coincidence worth handling carefully

Sorted by the system the owner named, the two categories nearest this subject sit a long way apart. The engine group is the largest in the whole file, with 200 filings carrying a recoverable odometer figure and a midpoint of 58,000 miles. Tracked as its own group, engine and engine cooling contributes 97 filings at a midpoint of 98,000. Take every group together and the file as a whole has a midpoint of 60,000.

Now lay those against the belt window published above, and notice the arrangement. The engine midpoint falls below the point at which that window even opens. The cooling midpoint falls well inside it, close to the top. Between them, the two engine-related groups in the archive straddle the entire interval this page is about.

Read strictly as sequence, it does support one honest conclusion. Across a good proportion of the cars behind those engine filings, the belt would still have been on its original fitment when the owner wrote in — the first replacement not yet due, let alone overdue. Engine trouble, in that file, tends to arrive before the belt window opens at all. For a buyer, that reframes what a car in the ordinary used-market mileage band represents: the belt service is very often still ahead of it rather than behind it, whatever else the folder shows. The broader version of that argument — what the same archive can and cannot tell you about a car’s remaining life — is worked through on our page about what actually counts as too many miles.

The recall record, and a silence that means something

The second archive is the safety recall file, and here the absence is the finding.

Across 145 distinct campaigns grouped into 24 component families, there is no timing belt family. There is no timing belt entry inside any other family either. The engine family itself holds just 5 campaigns across 3 manufacturers, and the archive names no parts for any of them — the campaigns are grouped at engine level with nothing more specific recorded.

That silence is neither an oversight nor a testimonial. A recall exists because a manufacturer has conceded a defect — something built wrong, specified wrong, or giving up earlier than the design said it should. A wear item arriving at the end of the life it was drawn for is the opposite of a defect. It is the design behaving exactly as intended. So a timing belt reaching its interval can never enter this record, however many engines it takes with it once it is ignored.

The practical consequence deserves stating plainly, because the recall check is genuinely one of the better free checks available to a buyer and this is precisely where it goes blind. Running a car’s number against the open-recall database is worth doing on every vehicle you consider. It will not, and structurally cannot, tell you anything about the single most engine-destroying item on the maintenance schedule. Two different questions, two different records, and a clean recall result answers only one of them.

Establishing whether it has been done

Here is the buying problem in one sentence. The component is invisible, its condition cannot be assessed, its interval is long enough that it usually falls to somebody other than the original owner, and the only evidence that the work happened is a document that a private seller has no obligation to have kept.

That makes this an unusual inspection. Almost everything else on a used car can be checked against the car. This one is checked against paper.

First, establish that the question applies

Before any of the paperwork matters, settle two facts about the specific vehicle: does this engine use a belt, and is it an interference design. The methods are set out earlier in this page and none of them takes long. Get those answers before the viewing rather than during it, because they determine how much weight everything else carries. A chain-driven engine turns this into a listening exercise. A belt-driven interference engine turns it into the most important line on your entire inspection.

What a genuine timing belt invoice looks like

A real one has a recognisable shape, and learning it takes a minute.

• The words, unambiguously. Timing belt, cam belt, or timing belt kit — naming the component, not a service tier. “Major service” on its own is a heading, not evidence.
• Parts listed individually. A belt with a part number. A tensioner. Idler pulleys, by quantity. Seals where they were done. If the engine has a belt-driven pump, a pump and a coolant line.
• A labour line proportionate to the job. Belt work carries hours, not minutes. A trivial labour charge against a timing belt heading is the single clearest sign that the document describes something else.
• Both a date and an odometer reading. You need both, because the interval has both. A receipt with one and not the other only answers half the question.
• An identifiable shop. A name, an address, a phone number that still connects. A workshop will usually confirm from its own records that it did the work, against the registration or the VIN, and that is one of the few pieces of corroboration in this whole exercise that the seller does not control.

Two failure modes account for most of the confusion here, and both are usually innocent. The first is that a seller genuinely believes the timing belt was done because an accessory belt was fitted at some point, and the invoice, once you read it, says so — a modest part and a short labour line. The second is that a “full service” is remembered as having included everything. Neither is dishonesty. Both are resolved by reading the document instead of discussing it, and both are the reason to ask for the paperwork rather than to press the point verbally.

The method for reading a folder as a whole — the cadence of the receipts, the paper they are printed on, whether the mileages progress sensibly against the dates — is set out on our guide to how to read a service history properly, and every one of those tests applies here. One difference is worth flagging. Oil changes recur, so you are reading a pattern. A timing belt happens once or not at all in most ownerships, so you are searching an entire folder for a single event, and its absence is not a gap in a rhythm. It is the whole answer.

Corroboration on the car itself

None of this proves anything on its own. All of it is worth a glance, because it either supports the paperwork or quietly contradicts it.

• A service sticker. Many workshops fit a small printed label under the bonnet or on a slam panel with the date and mileage of the belt change. It is not proof, but it is a strong corroboration when it matches an invoice, and an interesting question when it does not.
• Disturbed fasteners on the timing covers. Bolt heads with clean marks on an otherwise dirty engine, or fresh sealant where an old cover has been off, mean somebody has been in there. It does not tell you when or what they did.
• A cover, mount or crank pulley that is conspicuously cleaner than everything around it. Same finding, same limits.
• Coolant that looks recently changed on an engine with a belt-driven pump. Consistent with a belt job having been done, and worth mentioning to the seller to see whether the story lines up.
• Noise from the timing end at a cold start. Insist on finding the car unstarted. A rattle or whine from the cover area belongs on the inspection brief regardless of what the folder says.

Where the belt is concerned, the honest limit is worth repeating: even if a cover can be opened and a stretch of belt exposed, you cannot judge remaining life from what you see. Any inspection of that kind is looking for contamination and obvious damage, not for wear. Which is why this item, more than most, belongs on a written brief for an independent pre-purchase inspection rather than being left to a look round the engine bay. Ask the inspector specifically whether the engine is belt or chain driven, whether it is interference, and what evidence exists that the belt has been done. Those are three answerable questions, and an inspector who does this for a living answers them faster and more reliably than you will.

When there is no record at all

This is the common case, and it deserves an unsentimental answer rather than a balanced one.

On most service items, missing paperwork widens a range. There is no separating a careful owner who binned the paperwork from one who never did the work, so pricing the uncertainty and moving on is a fair response. That reasoning does not transfer here, for one reason: the consequence is not symmetrical. Being wrong about oil changes costs you some engine life. Being wrong about a belt on an interference engine costs you the engine, at a moment of its choosing, possibly on the drive home.

So the working assumption on an undocumented belt is that it has not been done. The base rate supports it. Plenty of people sell a car before its belt window has ever opened, nothing on the dashboard reminds anybody the job is due, the car gives no symptom to prompt it, and it is expensive enough that owners who are not planning to keep the vehicle have an obvious incentive to leave it for whoever comes next. A seller who says it was done but has nothing to show for it is not necessarily lying — they may be repeating what they were told when they bought it, which is how an undone belt gets passed along a chain of owners each of whom believed somebody else had dealt with it.

The paperwork the seller keeps is also, by definition, the paperwork the seller chose to keep. The records attached to the vehicle itself sit outside that curation entirely, and a check of what is recorded against the VIN covers title history, reported readings and whether the odometer progression makes sense — which matters here more than usual, because every mileage-based argument on this page collapses if the mileage is not real.

Pricing it into the offer

An undocumented belt is not, by itself, grounds for abandoning a car. It is grounds for changing the number, and it converts into one cleanly, which makes it one of the more straightforward negotiating positions available to a buyer.

Do it in this order.

• Get a real quote for that engine, not a general figure. Call an independent workshop that knows the make and ask what a full belt service on that specific engine comes to, with the tensioner, the idlers, the seals and the pump if it is belt-driven. This takes one phone call and replaces a vague worry with a specific amount.
• Decide which of the three outcomes you want. Ask the seller to have the work done before the sale, with the invoice handed over. Ask for the price to come down by the quoted amount. Or accept the car as it is at a price that reflects the work being immediately due. The first is the cleanest, because the work is verified before your money moves.
• Say what the adjustment is for. A deduction attached to a named job with a named quote behind it is a different conversation from a general request for a discount, and it is much harder to argue with. Our guide to judging where a car sits in its life makes the same case for the other big-ticket items that cluster around this mileage band.
• Do not treat “probably fine for now” as a plan. If the belt is overdue on either the mileage or the years, every drive is the same drive, including the one home from the viewing. There is no low-risk period at the start of an overdue belt’s remaining life.

The framing that makes this easy to hold on to is the asymmetry. You are not haggling over a service. You are buying out an unbounded risk for a bounded price, and the bounded price is knowable to the pound before you commit. Very few decisions in a used-car purchase are that clean.

What to actually do

One last point of proportion, because this page has necessarily dwelt on the worst outcome. A belt service performed on time is an ordinary, predictable, entirely manageable cost of owning a car with that kind of engine, and millions of them are done every year without incident. The failure this page describes is not what happens to belts. It is what happens to belts that were never changed. The difference between those two situations is a piece of paper and a decision somebody made, which is exactly why a buyer who reads the folder properly is buying a different risk from one who does not.

And the wider caution applies here as everywhere. A belt is a cost you can see coming, quantify, and plan around. What you cannot see coming is whatever happened to this particular car in the years before you met it, and no amount of time spent at the front of an engine will surface any of it. That material lives in paperwork nobody at the viewing controls, so it is worth pulling a report on what the vehicle’s number has recorded against it before you agree a figure.

Common questions

What actually happens if a timing belt breaks?

It depends entirely on one design characteristic. On a non-interference engine the camshaft stops, the engine stops, and nothing inside is damaged — the car is recovered, a belt is fitted, and it drives away. On an interference engine the pistons and the valves share the same space at different moments, and when the belt goes they no longer take turns. The result ranges from bent valves at best to a damaged cylinder head, damaged pistons and a bent connecting rod at worst, with engine speed at the moment of failure deciding most of it. On a car of ordinary value that is frequently the point at which it is written off rather than repaired.

How do I find out whether my engine is interference or non-interference?

Not from the model name, because one model line is often sold with several engines and they need not agree. The reliable routes are the owner’s handbook and manufacturer schedule for that specific engine, a franchised dealer’s parts department resolving the VIN to the exact engine, or a specialist independent workshop that works on that make daily. Owner communities for the engine code are useful corroboration. If you cannot establish it, treat the engine as interference: the cost of being wrong that way is one unnecessary service, and the cost of being wrong the other way is the engine.

Is a timing chain really maintenance-free?

No, and it is worth resisting that framing when a listing offers it. A chain carries no scheduled replacement interval and rarely snaps outright, which is a genuine advantage. But chains wear at every pin and bushing and grow longer as a result, hydraulic tensioners eventually run out of travel, and the plastic guides wear grooves and crack. The tells are a rattle from the timing end in the first seconds of a cold start, a gradual loss of performance and economy, and a stored fault for the cam and crank sensors disagreeing. When a chain does need doing, the job is usually dearer than a belt job, because the chain lives inside the engine rather than behind a plastic cover.

Are there warning signs before a timing belt fails?

Usually not, and it is more useful to say so than to invent a symptom list. The belt is enclosed, it cannot be assessed by eye in any meaningful way, and a belt near the end of its life looks much like a new one. What you can sometimes catch is a bearing noise from the tensioner or an idler, oil or coolant leaking onto a belt that must run dry, or the rough running and sensor-correlation fault that follows a belt having already jumped a tooth. Every one of those is a warning about something around the belt rather than the belt itself. This is precisely why the item is on a schedule: a wear part that cannot be inspected has to be replaced on time instead.

Should the water pump be replaced at the same time?

Where the pump takes its drive from the timing belt, almost always, and the reason is geography rather than salesmanship. The pump sits behind the same covers and is exposed the moment the belt is off, so fitting one at that point adds a part to a job already being done. Leave it, and if it fails later the entire access job is repeated and the belt fitted alongside it is scrapped in the process, because a timing belt is never reused once its tension has been released. On an engine whose pump is driven by the accessory belt at the front instead, the argument does not apply and the two jobs are unrelated.

The seller says the belt was done but cannot find the receipt. What should I do?

Treat it as not done, without treating the seller as dishonest. This is exactly how an unchanged belt travels down a chain of owners: each one was told it had been dealt with and passed the same reassurance on. Ask which garage did it and when, and call them — a workshop will usually confirm from its own records against the registration or the VIN, and that is corroboration the seller does not control. If nothing can be confirmed, get a quote for a full belt service on that specific engine and put the figure into the negotiation. The cleanest outcome is the work being done before the sale with the invoice handed over.

Does a clean recall check tell me anything about the timing belt?

Nothing at all, and the reason is structural rather than accidental. Across the 145 campaigns in our archived recall pull, grouped into 24 component families, there is no timing belt family — and the engine family holds only 5 campaigns across 3 manufacturers, with no parts named. A recall records a defect the manufacturer has conceded. A wear item arriving at the end of the life it was drawn for is not a defect; it is the design doing what it said it would. So a belt at its interval can never appear in that record however much damage it goes on to do. Check for open recalls on every car you consider, and understand that the check is answering a different question from this one.