Transmission Repair Cost: Four Routes, Not One Job

Most repair bills are an irritation. This one is a fork in the road. A transmission is among the very small number of components on a car whose replacement can cost more than the car will fetch, which means the useful question is almost never how to fix it. It is whether to fix it, by which route, and — for anyone reading this in a car park rather than a waiting room — whether to take on a vehicle that may be heading this way at all.

Before any of the money, one piece of public evidence is worth putting on the table, because it says something about where transmission trouble comes from. This site keeps an archive of NHTSA safety recall campaigns for the models it holds records on, grouped by the component family the agency files each one under. Read that way, the power train family carries 10 campaigns. So does the electrical system. What separates them is how far those campaigns are spread: the electrical ones come from 7 different manufacturers, and the power train ones come from 3.

That is the shape worth looking at. Families sitting high and wide — the gasoline fuel system with 16 campaigns across 10 manufacturers, air bags with 16 across 8 — behave like problems the whole industry has, arriving at unrelated companies from shared suppliers and shared physics. The power train sits high and narrow. As many campaigns as the electrical system, from fewer than half as many companies.

Now the fence around that, because it needs one. This archive is not a census of American recalls and cannot be read as one. It holds 145 distinct campaigns across 24 component families, reduced from 158 returned rows once 13 duplicates were removed, and it was pulled per vehicle for the models this site covers rather than swept across the whole market. Nothing in it ranks manufacturers, and nothing in it identifies the worst transmission on the American road. The one comparison it will bear is an internal one: inside a single pull, some systems turn up at companies with nothing else in common, and some stay put.

Within the power train family, four of those campaigns name the automatic transmission itself. Two more concern the driveshaft, one the gear position indication, and one the shift lever and column linkage. So the concentration is not an artefact of the category being stuffed with unrelated driveline oddities — a real share of it is the gearbox. The reading a buyer should take from it is modest and specific: transmission defects serious enough to become a recall tend to belong to particular designs from particular companies, not to a bad decade for gearboxes generally. Which is an argument for researching the exact model and the exact unit fitted, and for checking whether a specific car has open recalls, rather than for adopting a rule about transmissions in the abstract.

Why there is no figure on this page

Somebody arriving at this page wants a number, and there is a version of this article that gives them one. It would be wrong in most of the country on the day it was published, and it would go on looking right for years.

The reason is not vagueness. It is that a transmission bill is a sum of parts that behave independently. The labour rate is regional and moves. The hours are set by the vehicle’s architecture, and the same nominal job can be half a day on one car and two days on another. The unit itself might be a soft-parts overhaul of the gearbox already fitted, or a remanufactured assembly with a national warranty, or a used one out of a breaker’s yard, and those are three different products at three different prices with three different risk profiles. Then there is what else has to be done while the thing is out, which nobody can quote until it is out.

A single range flattens all of that into a figure that reads as authoritative and is doing no work at all. Worse, it invites the wrong comparison. A reader carrying a remembered number into a workshop tends to argue about margin, when the honest disagreement is almost always about scope: which route, and what gets renewed while access is free.

So the useful thing this page can do is tell you what sets the number, which parts of it you can establish for a specific car before you commit, and how the answer interacts with what the vehicle is worth. Those are questions with real answers.

Four routes out, and what each one buys

Almost every conversation about a failed transmission is really a conversation about which of four things to do. They are frequently discussed as though they were price tiers of the same job. They are not — they differ in what is renewed, what is guaranteed, and how much of the unit’s history you inherit.

Repair in place

Some faults are bounded and do not require the unit to come out. External solenoids, a wiring harness or connector that has chafed, a speed sensor, a leaking pan gasket or cooler line, a shift cable out of adjustment, a failed transmission mount, or on many designs a valve body that can be reached from below with the pan off. On some vehicles the control module is external and can be replaced or recalibrated on its own.

This is the outcome worth chasing hard before accepting a larger one, because the gap between it and any of the other three routes is enormous. It is also the outcome that a hurried diagnosis skips past. A transmission that misbehaves in one gear, or that throws a code and clears, or that shifts badly only when cold, is a candidate for a bounded repair. A transmission that slips under load is not.

Rebuild the unit that is in the car

A rebuild means removing the transmission, dismantling it, cleaning everything, replacing the soft parts as a matter of course — friction plates, steels, seals, gaskets, bushings, sometimes the pump and the converter — and replacing whichever hard parts are found damaged. The case, and usually the major shafts and gearsets, are the ones already in the car.

What a rebuild buys you is control and, often, an improvement. A specialist who knows a given transmission family knows its weak points and can fit uprated parts where the original design was marginal. What it costs you is variability. The final scope is not known until the unit is apart, so the estimate is an estimate. The warranty is the shop’s own, which makes the shop’s reputation and durability part of the product. And the quality range across rebuilders is genuinely wide, in a way that is invisible from the forecourt.

Fit a remanufactured unit

A remanufactured transmission is built to a specification in a facility set up to do nothing else, from cores gathered in volume. Every unit gets the same defined list of replaced components, whether or not that particular core needed them, plus any updates the manufacturer issued after the original design went into production. The unit is dynamometer-tested before it leaves.

The practical advantages are consistency, a warranty that is usually national and usually longer than a local shop can offer, and speed — a unit arrives rather than being built while your car occupies a lift. The disadvantage is that you have no relationship with whoever built it, and the warranty comes with conditions that matter: typically it must be installed by a shop rather than a driveway, the cooler must be replaced or properly flushed, and the fluid specification is not negotiable. Read those conditions before the unit is ordered, not after it fails.

Fit a used unit

A used transmission from a dismantled vehicle is the cheapest part on the list and the largest gamble. You are buying an assembly whose service history is unknowable, whose mileage is whatever the donor’s odometer claimed, and which has been sitting in a yard. Guarantees exist but are usually short, usually parts-only, and frequently exclude the labour to fit it — which is where most of the money is.

There are situations where it is the correct answer. A car near the end of its economic life where the alternative is scrapping it. A donor with documented low mileage and a verifiable history. A well-understood, robust transmission family where the failure mode being replaced is rare. Outside those, the arithmetic is unkind, and the reason is the next section.

The part of the bill that does not change

Here is the fact that reorganises the whole decision, and it is the one most often left out of the conversation: the labour to get a transmission out of a car and back into it is very nearly a fixed cost. It does not care which of the four units you chose.

Consider what that labour involves on a typical vehicle. The fluid comes out. The battery and airbox come off for access. Cooler lines are disconnected and plugged. Driveshafts or the propshaft come out, which on a front-drive car means separating suspension joints. Electrical connectors, the shift linkage or cable, and the speedometer or output sensors are released. On many designs the exhaust has to come down. The torque converter is unbolted from the flexplate through a small aperture, one bolt at a time, rotating the engine between each. The engine is supported from above or below because the transmission is a structural member. On a great many transverse front-drive cars the entire subframe drops, which disturbs the alignment and adds a wheel alignment to the bill on the way out. Then the unit comes down onto a jack, and every step is reversed.

Some vehicles are worse. On certain designs the engine and transmission come out as an assembly, either upwards or on the subframe, because there is no other way. On others the body has to be lifted from the chassis. None of that is exotic. It is ordinary, and it is decided by the car rather than by anybody’s pricing policy, which is why two vehicles with much the same symptom can produce quotations that look nothing alike.

Three consequences follow, and all three are counter-intuitive.

• The cheap unit saves less than its price suggests. If a large share of the total is removal and refitting, choosing the cheapest assembly reduces the total by a much smaller proportion than it reduces the parts line. People choose the used unit imagining they are halving the bill, and are surprised by the invoice.
• Doing the job twice is the disaster case. A used unit that fails inside a year does not cost you the price of a second used unit. It costs you the whole removal and refitting labour again, on top of what you already spent, and you are then making the same decision from a worse position. This single risk is why remanufactured units with real warranties often win on expected cost even when they lose on sticker.
• Everything adjacent should be done while it is out. The rear main seal, the flexplate, the torque converter, the transmission mounts, the fluid cooler and its lines, and on many cars the starter, are all trivially accessible with the gearbox out and painfully expensive to reach later. Deferring any of them to save a little now is buying a second removal at full price.

Why a CVT is a different, and usually worse, proposition

The servicing side of this — the fluid a CVT needs and why it is not interchangeable with anything else — belongs to our guide on how often to change transmission fluid, and there is no point repeating it here. What that page does not cover, and what matters once something has actually gone wrong, is that a CVT changes which of the four routes above remain open to you.

Start with where the wear ends up. In a conventional automatic, the components that give out first are largely the ones a rebuild replaces as standard: friction material, seals, bushings, the pump. That is precisely why the rebuild trade exists and why kits are sold for it — the consumable parts and the core parts are different parts. A CVT does not divide so conveniently. Its wear concentrates on the surfaces that transmit drive and on the element running against them, and those are core components, not consumables. Once those surfaces are marked, restoring the unit means renewing the expensive half of it, which is most of the reason there is no thriving CVT overhaul trade to walk into.

Two further things narrow it. Fewer independent specialists have the tooling, the training or the volume to take a CVT apart with confidence, so in many areas the local rebuild option simply is not offered. And the parts supply behind an overhaul is thinner, because the aftermarket has had less reason to develop one.

What that leaves, in practice, is remanufactured or used — the two routes with the least flexibility on price. The repair-in-place route survives for genuinely peripheral faults, but it also gets used up faster, because the symptom set a CVT produces when it is unhappy is less diagnostic than an automatic’s. A conventional automatic that misbehaves in one gear has told you something specific. A CVT has no gears to isolate, so shudder, drone and slip resolve to the assembly far more often than they resolve to a component.

There is a modern complication worth naming too. Many CVTs are programmed to imitate fixed ratios under hard acceleration, partly because buyers dislike the sensation of a constant engine note. That programming can mask early slip, because the behaviour a driver would notice as wrong is being deliberately produced by the software. On a test drive it means the honest information is at light throttle and low speed, not at full throttle.

What a transmission does before it fails

Transmissions rarely fail without warning. They fail after a period of warning that was interpreted as a quirk. The value in knowing the sequence is not that it lets you fix anything — it is that it tells you which of the four routes you are heading toward, and it gives a buyer a reason to walk away before the problem is theirs.

The most useful division is between symptoms that commonly resolve to something bounded and symptoms that commonly resolve to internal wear.

Symptoms that are often something smaller

• One gear or one shift misbehaving while everything else is normal. That pattern points at a specific circuit — a solenoid, a valve, a clutch pack fed by one passage — rather than at general wear.
• A fault that is present cold and gone warm, or the reverse. Temperature-dependent behaviour frequently comes from clearances, seals or an electrical fault that changes with heat, and these are among the more repairable causes.
• A fine vibration at steady light throttle at cruising speed, felt as though the road surface changed. This is often the torque converter lock-up clutch rather than the gearbox proper, and on some vehicles it has a known fix.
• Limp mode that appears once, clears on a restart and does not return under normal load. Worth scanning and worth watching, but a single event is not a diagnosis.
• Harsh engagement with no other symptom, which is a broken or collapsed transmission or engine mount often enough to be worth checking before anything expensive is contemplated.

Symptoms that usually mean the unit

• Slip under load. The engine revs rise and road speed does not follow, most obviously accelerating uphill or overtaking. Watch the tachometer rather than trusting the feel. This is the one symptom on the list that almost never has a cheap explanation.
• A long delay selecting drive or reverse, particularly one that gets worse as the car warms up. A pause of a second or more before the car takes up drive is a finding, not a characteristic.
• A gear that has gone missing, or no reverse at all.
• Noise in neutral that changes when a gear is selected. Noise that is present with the car stationary and the transmission in neutral implicates the input side — bearings, the pump — and it is inside the case.
• Repeated limp mode that returns immediately under load, or a car that will only offer one fixed ratio.
• Overheat warnings on a car that tows or climbs, especially if they now appear on journeys that used to be uneventful.

Fluid condition sits alongside all of this and is dealt with properly on the servicing page — the short version being that a burnt smell is a finding about clutch material rather than about fluid, and that metallic glitter is a finding about hard parts.

One more caution, which applies in both directions. The transmission is the most frequently mis-blamed assembly on a car. An engine misfire under load reads as a shudder. A worn constant-velocity joint or a failing wheel bearing reads as a driveline noise. A dragging brake reads as a car that will not pull away properly. Before accepting a transmission diagnosis on a car you are being asked to pay for, ask what was ruled out and how.

When power train trouble actually surfaces

There is a second dataset worth bringing to this, and like the recall archive it needs describing accurately before it is used.

The National Highway Traffic Safety Administration collects complaints from owners who believe something on their car is dangerous, and its intake sentence sometimes records what the odometer read that day. Out of 9,096 complaints in the archive, 1,324 gave up a usable odometer figure — a coverage rate of 14.6 percent — drawn from 13 models, 15 component categories and three model years, 2015, 2016 and 2018. Filed under power train there are 124 such readings. Their midpoint is 49,000 miles. The midpoint for the whole file is 60,000.

Rank, then. Power train ties with the electrical system for the earliest position in the file, both at 49,000 miles. The engine comes next at 58,000. Two categories sit conspicuously late by comparison: gasoline fuel system at 89,000, and engine and engine cooling at 98,000, a long way behind the power train. So among the faults owners judged dangerous enough to report, the transmission is not a late-life problem. It arrives early, and it arrives ahead of the 60,000-mile midpoint of the archive it belongs to.

Most buyers carry a mental model that puts transmission trouble deep into six figures and safely inside somebody else’s ownership. These filings do not support it. They cluster in exactly the mileage band that used cars are advertised at, which is the band you will be shopping in. It is one reason the link between an odometer reading and real risk is weaker than it appears, a point taken apart at length in our guide to how many miles is too many, which draws on the same archive.

The repair set against what the car is worth

Now the decision this page exists for. Almost every component on a car can be replaced for less than the car is worth. A transmission is one of the few where that is not reliably true, and once it stops being true the arithmetic changes character entirely.

The instinct at this point is to compare the quotation against the car’s value, decide the repair is more than the car is worth, and walk away. That instinct is right about half the time and wrong the other half, because it is the wrong comparison. Here is the right one.

The comparison that actually works

Ask what this specific car, with a healthy transmission in it, would cost you to replace with an equivalent vehicle — the same age, the same condition, the same mileage, bought at retail with all the fees and taxes that come with a purchase. Then set the repair against that, not against the trade value of the broken car. The trade value of a car that does not drive is close to scrap, so measuring against it makes every repair look extravagant.

Then adjust for four things.

• What else the car needs. A transmission decision on a vehicle that also wants tyres, an exhaust, suspension work and has structural corrosion is a different decision from the same quotation on a car that is otherwise sound. Price the next two years, not the next repair.
• What you know about this car and cannot know about the next one. You have its history, its habits, its service record. The replacement is an unknown with an unknown transmission in it, bought at retail. This factor argues for repair more often than people credit, and it is the one most consistently left out.
• What the repair leaves you holding. A remanufactured unit with a real warranty does not just fix the car — it converts the single largest unbounded risk in owning it into a bounded one with a term attached. On a car you intend to keep, that is worth more than the price gap between routes.
• What the repair does not do. It does not add its own cost to the car’s resale value. Fitting a rebuilt transmission to an old car makes it a car that drives, not a car worth what it was worth plus the invoice. If the plan is to repair and immediately sell, expect to lose most of it.

When walking away is right

The case for abandoning the car is strongest when several of these are true at once: the vehicle is at an age and mileage where other large systems are also near the end of their lives; the transmission is one where only the expensive routes are open; there is corrosion, which is the one fault that ends cars regardless of what else is right; and the money required is a meaningful fraction of the cost of a sound replacement. Any one of those alone is not decisive. Three together usually are.

And be clear-eyed about sunk costs. Money already spent on diagnosis, or on the car generally, is gone whichever way you decide. The only question that matters is which of the two available futures costs less from here.

The buyer’s section: testing a transmission at a viewing

Everything above matters more before the purchase than after it. A transmission is the single most expensive thing to be wrong about on a used car, and a careful half hour will catch a good proportion of the trouble.

The most important condition is one that costs nothing and that sellers routinely defeat: drive the car from genuinely cold. Ask for it not to be started before you arrive, and check the bonnet with the back of your hand when you get there. A warm engine on a car that was supposedly sitting is information in itself. Cold is when delayed engagement, flare between ratios and a shudder that disappears once everything is warm are all visible, and a pre-warmed car has quietly removed your best test.

Then work through it in order.

• Stationary, engine idling, foot firmly on the brake. Select reverse, then drive, then back again, leaving a couple of seconds between each. Engagement should be immediate and gentle. A pause, a clunk, or a rise in revs before the car takes up drive is worth writing down.
• Reverse up a slight incline if there is one, then hold the car on the brake and let it creep. Reverse loads different components from drive on many designs, and a fault that only shows in reverse is common.
• Drive gently through the lower ratios from cold. Note whether shifts arrive when expected and whether any of them thump, flare or hunt between two ratios at a constant speed.
• Then open the throttle firmly to force a downshift and a hard upshift. Watch the tachometer. Revs that climb without matching road speed are slip.
• Then a steady, light-throttle cruise at whatever speed the car settles into top gear, and hold it. A fine, cyclical vibration here is the lock-up clutch. On a CVT, this is also where drone or a low-speed shudder is honest, because the software is not busy pretending to be a stepped gearbox.
• Lift off and let it come back down the ratios. Overrun downshifts should be smooth and unhurried.
• Afterwards, get your head under the car and look for anything freshly wet around the sump, the joint with the engine and the metal lines running out to the cooler. Then check whether the dashboard has acquired any warning it did not have on the first start.
• Have it scanned. The transmission control module keeps its own history, and stored codes outlive most sellers’ expectations. Where the tool supports it, adaptation values and internal counters say more than the codes do.

Two of those steps want a lift and a scan tool, which is the argument for writing a pre-purchase inspection into the deal on any car where the gearbox is the thing you would least like to have misjudged. What no amount of driving will tell you is the paper side of the vehicle: whether the odometer figure is honest, whether the car was declared a total loss and put back together, whether it stood in floodwater. A VIN history check settles those quickly, and it matters more here than on most subjects. A false mileage figure does not merely mislead you about the car; it invalidates every judgement on this page, since all of them are measurements taken against the odometer.

What a recently replaced transmission actually means

Sooner or later you will see a listing that mentions a new or rebuilt transmission, framed as a selling point. It might well be one. It is also the single most information-dense claim in the advertisement, and it deserves four questions.

Which route was taken? “New transmission” almost never means new. It means rebuilt, remanufactured or used, and those are three very different purchases. A remanufactured unit from a known builder with paperwork is a genuinely good outcome. A used unit from a breaker, fitted by whoever was cheapest, has moved the risk around rather than removed it.

Who did the work, and is there an invoice? A receipt naming the shop, the unit, its part or serial number and the date is the difference between a fact and a claim. Ask for it. A seller who has it will produce it happily, because it cost them a great deal of money and they know it is worth something.

Is there a warranty, and does it transfer? Many remanufactured units carry a term that survives the sale, and some do not. Some require the warranty to be registered, or the servicing to be documented, to remain valid. This is worth establishing precisely, because a transferable warranty on the most expensive component in the car is a real asset and should be reflected in what you are willing to pay.

Was the cause fixed, or only the symptom? This is the one that catches people. Transmissions rarely fail for no reason. If the fluid cooler — which on many cars lives inside the radiator — had failed and let coolant into the transmission, and the radiator was not replaced, the new unit is drinking the same coolant. If debris from the old failure was left in the cooler and the lines, it is circulating through the replacement now. If the car overheated because it was towing beyond its rating, it is still towing beyond its rating. A replacement fitted without addressing why the first one died is a repair with a countdown attached, and this is by some distance the most common reason a replacement transmission fails early.

There is also a subtler point about mileage. Once a transmission has been changed, the unit’s own history and the odometer reading have parted company, in both directions. A car showing high mileage with a recent remanufactured gearbox may be a better proposition than an identical one showing less with its original unit and no service history at all. Conversely, a used transmission out of an unknown donor means the odometer on the dashboard says nothing about the wear in the most expensive assembly under the car.

Finally, treat it as a signal about how the vehicle was used. A transmission replaced early in a car’s life points either at a design with a known problem — worth checking against the recall record, as above — or at duty the vehicle was never really meant for. Neither is disqualifying. Both are worth knowing, and prior use of that kind, including fleet, rental and commercial registration, is one of the things a records search against the VIN can surface when the seller has no way of telling you.

What to do with all this

Common questions

Is it worth repairing a transmission, or should I just buy another car?

Compare the repair against what it would cost to buy an equivalent car at retail, including fees and taxes, rather than against what your car would fetch in its current non-driving state. Then weigh what else the vehicle needs over the next couple of years, whether it has corrosion, and how much you know about its history that you would not know about a replacement. The case for walking away is strongest when the car is at an age where other large systems are also due, when only the expensive repair routes are open, and when the money involved is a meaningful share of a sound replacement. One of those on its own rarely decides it.

What is the difference between a rebuilt and a remanufactured transmission?

A rebuild uses the unit already in your car: it is dismantled, the soft parts are replaced as standard, and damaged hard parts are replaced as found. The scope is not fully known until it is open, and the warranty is the rebuilding shop’s own. A remanufactured unit is built in a dedicated facility to a fixed specification from a pool of cores, with the same defined component list replaced every time plus any updates issued after the original design, and it is tested before dispatch. Remanufactured units generally carry longer and more portable warranties; rebuilds offer more control and the option of uprated parts where a design was known to be marginal.

Is a used transmission from a breaker a bad idea?

It is the cheapest part and the largest gamble, and the reason is that the labour to fit it is the same as the labour to fit anything else. If it fails, you pay that labour a second time, which frequently costs more than choosing a better unit would have. It can still be the right answer on a car near the end of its economic life, on a donor with a documented history, or where the failure being replaced is rare in that transmission family. Check what the guarantee actually covers — most are short and cover parts only, which excludes the expensive half of the job.

Are CVTs really more expensive to repair than conventional automatics?

The problem is less about the price of any one route than about how few routes remain open. In a conventional automatic the parts that wear first are largely the ones a rebuild replaces as standard, which is why an established overhaul trade exists for them. In a CVT the wear concentrates on core components rather than consumables, comparatively few independents will take one on, and the parts supply behind an overhaul is thinner. What is left is usually remanufactured or used, so the gap between a bounded repair and a full replacement is narrower and the point at which the work exceeds the car’s value arrives sooner.

Can a transmission problem turn out to be something cheap?

Yes, and it is worth ruling that out before accepting anything larger. Solenoids, speed sensors, a chafed harness or connector, a leaking cooler line, a maladjusted shift cable, a collapsed transmission mount, a valve body reachable with the pan off, and on some vehicles a control module recalibration are all bounded repairs that do not require the unit to come out. The pattern that points that way is a fault confined to one gear or one shift, a fault that changes clearly with temperature, or a warning that appears once and does not return under load. Slip under load, where the revs rise and the car does not, is the symptom that almost never has a cheap explanation.

Does NHTSA data show which transmissions fail?

Not in the way most people want it to. The recall archive on this site holds 145 campaigns across 24 component families, pulled per vehicle for the models covered here rather than across the whole market, so it supports no claim about which transmission is worst. What it does show is that power train campaigns are unusually concentrated — 10 campaigns from just 3 manufacturers, against 10 electrical campaigns from 7 — which suggests specific designs rather than an industry-wide weakness. The complaint archive is separate: 124 power train readings with a median odometer of 49,000 miles, which records when owners reported something dangerous, not when transmissions wear out.

Should I buy a used car that has had its transmission replaced?

Possibly, and on the right paperwork it can be the better of two similar cars. Establish which route was taken, because rebuilt, remanufactured and used are three different purchases and only one of them normally carries a long transferable warranty. Ask for the invoice naming the unit and the shop. Ask what caused the original failure and what was done about it — a replacement fitted without replacing or flushing the cooler, or without fixing whatever overheated the first unit, is on a countdown. And remember that the odometer reading no longer describes the transmission, in either direction.