How Many Repairs Make a Lemon? The Number That Actually Matters.
Ask someone how many repairs make a car a lemon and you'll hear a confident number. Three. Or four. Or "thirty days in the shop." The confident number is usually half-right and half-wrong, because the real test isn't a magic count, it's whether the manufacturer got a reasonable number of attempts to fix the same substantial defect and failed. There's a presumption that turns specific counts into a legal shortcut, but the heart of every lemon claim is proof: proof of the defect, proof of the attempts, proof that the same problem kept coming back.
And here's the thing that decides most cases before a lawyer is ever involved: that proof is built at the service counter, months before you know you have a claim. So let me walk through what "reasonable repair attempts" actually means, the presumption that helps you, the "substantial defect" standard people misread, and the documentation habit that wins.
"Reasonable" is the standard; the count is the shortcut
The underlying legal test in most lemon laws is that the manufacturer had a reasonable number of opportunities to repair a substantial defect and couldn't. "Reasonable" is deliberately flexible, because a dangerous brake failure shouldn't get as many tries as a balky window. What's reasonable for a defect that could kill you is fewer attempts than for a defect that's merely a serious nuisance.
To make that flexible standard usable, many states layer on a presumption, a set of specific conditions that, if met, presume the manufacturer had a reasonable number of attempts. California's version, drawn from the Tanner Consumer Protection Act, is the model most people have heard pieces of. It generally presumes a lemon if, within the first 18 months or 18,000 miles, one of these happened: the manufacturer made two or more attempts to repair a defect likely to cause death or serious injury; or four or more attempts to repair the same substantial defect; or the vehicle was out of service for repairs for a cumulative total of more than 30 days. Hit one of those, and the law presumes you've given the manufacturer enough chances.
Here's the part people miss in both directions. The presumption is a shortcut, not the only path. If you meet it, you get a strong legal leg up. But you can still have a valid lemon claim outside the presumption, more than four attempts after the 18-month window, say, or a defect that took five tries instead of four, it's just that you don't get the automatic presumption and have to prove "reasonableness" more directly. So a car that doesn't tick the presumption boxes isn't automatically not a lemon. And the specific numbers vary by state, so California's two-four-thirty isn't universal. The principle, reasonable attempts at the same substantial defect, is what's universal.
What counts as a "substantial defect"
The other half of the test is the defect itself, and this is where claims quietly fall apart. The problem has to be a substantial defect, one that significantly impairs the vehicle's use, value, or safety. Not every annoyance qualifies.
A defect that affects safety, brakes, steering, the car stalling in traffic, airbags, is squarely substantial. A defect that makes the car significantly less usable, a transmission that won't shift right, an engine that won't reliably start, persistent electrical failures that disable functions, generally qualifies too. A defect that materially cuts the car's value can count. What usually doesn't qualify: cosmetic issues, minor rattles, a trim piece that won't stay clipped, things that are irritating but don't meaningfully impair use, value, or safety.
There's also a "same defect" thread that matters. The repeated attempts generally need to be for the same underlying problem, not four unrelated visits for four different things. A car that went in once for the transmission, once for a window, once for the radio, and once for a tire isn't presenting one substantial defect that resisted repair; it's presenting four separate minor issues. What builds a lemon claim is the same significant problem coming back, again and again, despite the shop's attempts to fix it. That recurrence is the spine of the case.
For modern cars, especially electric vehicles where software and battery issues blur the picture, the "same defect" question can get genuinely tricky, the manufacturer may argue each software update addressed a "different" issue, when from your seat it's the same malfunction recurring. Framing those recurrences as one persistent substantial defect, rather than a string of unrelated tweaks, is part of the proof.
The documentation habit that wins cases
Now the most important practical point in this entire piece. Lemon cases are won and lost on documentation, and the documentation gets created at the moment of repair, long before anyone's thinking about a claim. The single best thing you can do, starting with the very first time something goes wrong, is get and keep a written repair order for every single visit.
A proper repair order does several things at once. It records the date, which is what proves your attempts fell within the time-and-mileage window for the presumption. It records the mileage, which feeds both the window and the buyback's mileage offset. It records what you complained about, which is what proves the recurring defect. And it records what the shop did, or didn't do, which is what proves the attempt was made and failed. Without that paper, you're left arguing from memory against a manufacturer with lawyers, and memory loses.
The classic mistake is the casual fix. You mention the problem, the service writer says they'll "take a look" or "reset it," you leave without paperwork, and that visit, a real repair attempt, simply doesn't exist in the record. Insist on a written repair order every time, even when the shop waves the problem off as minor or says they couldn't reproduce it. Especially then, in fact, because a defect the shop "couldn't reproduce" but you keep experiencing is a documentation battle waiting to happen, and a written record of you reporting it each time is your ammunition.
Keep your own parallel record too. Note the date each time the problem occurs, not just when you bring it in. Save the rental and towing receipts. Photograph or video the malfunction if you can capture it. Keep every communication with the dealer and manufacturer. The car that "won't act up at the shop" is a common and maddening situation, and your own contemporaneous record is often what bridges the gap between your experience and the service department's shrug.
The days-out-of-service path
Don't overlook the cumulative-days route, because it catches cars that don't fit the repeat-attempt pattern. In many states, a vehicle that's been in the shop for repairs for a cumulative number of days, 30 in the California presumption, can qualify even if it wasn't a high number of separate attempts.
This matters for the car with one stubborn, complicated defect that the shop keeps for long stretches. Three weeks for a transmission, then another two weeks when it fails again, and you're past 30 cumulative days without four separate visits. The repair orders that document each stay, with in and out dates, are what add up to that total, which is one more reason the dated paperwork is everything. A loaner you've had longer than your own car is often a sign you're approaching this threshold.
Build the proof before you need it
The frustrating truth about lemon claims is that the work that wins them happens before you know you'll need it. By the time you're sure the car is a lemon and ready to make a claim, the proof either exists in your file or it doesn't, and you can't go back and create repair orders for visits that left no paper.
So treat it as a habit from the first sign of trouble. Get a written repair order every visit, with the date, mileage, your stated complaint, and the work performed. Keep them together. Track the recurring defect so you can show it's the same substantial problem, not a scatter of minor ones. Save receipts for everything the defect cost you. And know your own state's presumption numbers, because they tell you when you've crossed from "frustrating car" into "presumed lemon."
Do that, and when the moment comes, you hand a lawyer a clean file that proves the defect, the attempts, and the failure, which is the entire case. The owners who struggle are almost always the ones whose repair history lives in their memory instead of on paper. The number that makes a lemon matters, but the documentation that proves you reached it matters more.