When should you replace a garage door?
The diagnosis that comes first: which of the four assemblies failed, and whether the door is structurally finished at all.
The diagnosis guideDeciding · Riverside County
Assume a technician has been out and told you both are possible. That is the moment this page is for — and “which is cheaper today” turns out to be the wrong way to settle it.

Compare what each option buys per year of service, not what each costs today. A repair buys time on one assembly of a door that has four — the sections, the counterbalance, the operator, and the track and hardware. A replacement resets all four clocks at once and lets you choose the specification. So the decisive question is usually not the fault in front of you: it is how old the other three assemblies are, because on most doors everything was installed on the same day and will reach the end of its life at roughly the same time. Repairing one assembly on a door whose other three are equally worn means being back, and paying the visit again. Two further things move the answer: whether your door’s profile is still manufactured, and whether the opener has to be brought into line with California’s battery-backup rule as part of the job. Get both options quoted at the same visit — a repair figure and a replacement figure from one person who has seen the door is the only genuinely comparable pair you will get.
Call (951) 232-9778 Monday–Saturday, 8am–6pm · Closed Sunday
What this guide covers
This page assumes both options are on the table. If you are not yet sure whether the door is genuinely finished, the diagnosis is a different question and it has its own guide — the first link below.
Reframing it
It compares two numbers that are not measuring the same thing, which is how people end up doing the expensive thing twice.
A repair figure and a replacement figure are not comparable, because they buy different quantities of time. A repair buys you the remaining life of everything you did not replace. A new door resets everything and starts a fresh clock on all of it.
The comparison that actually works is cost per year of service. Not a figure we are going to publish — this site publishes none, and the honest reason is that neither number exists until somebody has measured your opening. But the arithmetic is yours to do once you have both quotes, and it is simple: divide each option by the number of years you can reasonably expect it to last, and compare those.
That is where the decision usually resolves, because the two options often look close on the invoice and are nowhere near each other per year.
The input that dominates it is one thing: how old are the other three assemblies? A garage door is four systems on four clocks, and on the overwhelming majority of doors they were all installed on the same day. Replacing a spring on a door whose rollers, track and opener are the same age buys you the interval until the next of them goes — which may not be long. The four-clocks model is set out in full here.
None of that makes replacement the right answer. Very often it does not: a sound door with one failed part and years left in the rest is a repair, plainly. The point is that the calculation is about remaining life across the whole assembly rather than about the fault in front of you.
We will not quote either option before somebody has seen the door, and the reason is on our pricing page rather than repeated here. What we will do is quote both at the same visit, which is the part that makes the comparison possible.
The reasoning error
It is the most common sentence we hear on this decision, and it points in exactly the wrong direction.
Money already spent on a door is gone whichever option you now choose. It cannot be recovered by repairing again, and it is not wasted by replacing. It has no bearing on which option is better from today — and yet it is the reason a great many doors get patched a fourth and fifth time.
The mirror image is just as common and just as wrong: “I only replaced the springs last year, so I should get my money’s worth.” The springs were worth their money for the year of service they gave. What they cannot do is make the sections, the track or the opener any younger.
The clean way to think about it is to ask what you would do if you had just moved into the house today and knew everything you now know about this door. That question strips out the history and leaves the actual decision, and people usually answer it quite differently from the one they were agonising over.
There is one honest exception, and it is a practical rather than an emotional one: recent work can genuinely change the arithmetic, because it is remaining life you have already paid for. A door with new springs and new rollers has two of its four clocks reset, and that is a real asset. The distinction is between the value the work still holds, which counts, and the money it cost, which does not.
Common situations
Most repair-or-replace questions are not close once the whole assembly is considered. These are the ones we are asked about most.
| The situation | Usually | Why |
|---|---|---|
| One failed part, everything else sound and not old | Repair, clearly | Three clocks have plenty left. Replacing the door buys remaining life you already have. |
| A failed part on a door where everything is equally old | Genuinely close | This is the real decision, and it is where cost per remaining year does the work. Get both quoted. |
| One damaged section, profile still manufactured, colour matches | Repair | A section swap on a sound door is straightforward and the result is invisible from the street. |
| One damaged section, profile discontinued | Usually replace | A mismatched panel on the front of a house is generally a worse outcome than a new door, and it cannot be undone later. |
| Sections sound, but the opener has also died | Repair both, usually | Two assemblies rather than four. But check the opener rule below — it may be less optional than you think. |
| Third or fourth unrelated failure inside a year | Replace | You are replacing the door one component at a time in the least efficient order, paying a visit charge each time. |
| Sections structurally gone — stiles, delamination, corrosion through | Replace, no real choice | There is nothing sound left to attach hardware to. This is the case where repair is not an option rather than a worse one. |
| Door works, you dislike how it looks | Your call, and legitimate | An appearance decision is a perfectly good reason. Just make it knowing the door is sound rather than being told it is not. |
| Selling within the year | Usually repair to sound and safe | See the section below — the calculation changes when you are not the one who gets the remaining life. |
Only two rows in that table are genuinely close. That is typical: most of these questions resolve quickly once somebody has looked at the whole door rather than the broken part — which is what the diagnosis guide is for.
The mechanics
This is the part nobody explains, and it is where most of the confusion in this decision comes from.
One person, one visit, both options in writing. That is the only genuinely comparable pair you will get, because both figures come from the same assessment of the same door by somebody who has seen it.
Quotes from separate companies for separate options are not comparable at all — they are two people describing two different jobs after forming two different opinions about the door’s condition. Comparing them tells you almost nothing.
We quote both as a matter of course when both are viable, and we say which we would do and why. You are welcome to disagree with the recommendation; what you should not have to do is guess at the alternative.
The repair quote: exactly which parts, the spring wire size and length where springs are involved, whether both springs are included on a two-spring door, and what is not being addressed. That last item is the one that matters for this decision — it tells you what is still on an old clock.
The replacement quote: the door construction and layer count, the finish, the hardware and track, the spring cycle rating, the seal, and whether the operator is included.
Both: parts and labour separately, warranty terms on each, and a total that does not move once work starts. The specification guide has the full list of what to ask on a new door.
“What would you expect to fail next on this door, and roughly when?”
Somebody who has genuinely assessed the whole assembly can answer that immediately, because they looked at all four clocks rather than at the broken one. Somebody who has looked only at the fault in front of them cannot, and the answer will be vague.
It is not a trick question and it is not adversarial — it is the single most useful thing you can ask, because the answer is the repair-or-replace decision. If the honest answer is “nothing for years”, repair. If it is a list, you have your reason to replace.
Beyond the door itself
Once both options are on the table, three considerations change the calculation more than the immediate failure does, and two of them catch people out.
This decides most damaged-section questions on its own, and age is a poor guide to it — some fifteen-year-old profiles are still current products and some eight-year-old ones are long gone.
If the profile is current, a section swap is straightforward and the result is invisible. If it is discontinued, you are choosing between a visibly mismatched panel on the largest surface at the front of your house and a new door. Establish this before anything else, because it can end the discussion immediately.
This is the one that surprises people at the worst moment. California has required battery backup on residential garage door openers since July 2019, and a replacement door may not be connected to an existing opener that does not have it.
So on the replacement side of your comparison, the opener may not be optional. If yours predates the rule, it belongs in the replacement quote — and a quote that omits it is not comparable to one that includes it. The rule and how to check your own unit has its own guide.
Note the asymmetry: this affects only the replacement option. Repairing an existing door does not trigger it.
If you are selling within the year, the arithmetic inverts. Cost per remaining year is a poor guide when the remaining years are somebody else’s, and the relevant questions become whether the door is safe, whether it works, and how it looks from the kerb.
In that situation the usual answer is to repair it to sound and safe rather than to replace it — with the honest caveat that on most tract elevations here the garage door is the largest single surface a buyer sees, so a genuinely tired door is doing visible work against the house.
Whether that is worth addressing is a conversation for your agent rather than for us. What we can tell you is whether the door is sound, which is the part that turns up in an inspection report.
None of the three is about the part that failed, which is the point of this page. The fault gets you the visit; these decide the outcome.
Questions
Six questions this comes up as — answered on method, since we publish no figures.
Ring and ask. There is no charge for a question and Beau has been doing this since 1996.
(951) 232-9778Mon–Sat 8am–6pmOn the day, repairing is nearly always the smaller figure. That is not the useful comparison, because the two options buy different amounts of time: a repair buys the remaining life of everything you did not replace, and a new door resets all of it.
Compare cost per year of service instead, once you have both quotes. The input that dominates it is how old the other three assemblies are — on most doors everything went on at once, so repairing one part of a worn assembly means being back.
Because neither number exists until somebody has measured the opening and seen the door. Headroom, sideroom, whether the opening is square, the condition of the jamb and what the existing operator can carry all change the figure, and none can be assessed from a description.
What we will do is quote both options at the same visit, in writing, so you have a genuinely comparable pair. The measure is free whether or not you go ahead with either.
It counts, but not for the reason people usually give. The money spent is gone either way and is not an argument for anything. What counts is the remaining life you now own: new springs mean one of the four clocks is reset, and that is a real asset.
The question is still what the other three have left. New springs on a door whose sections are failing have not changed the sections.
You largely cannot, and it is worth knowing that rather than trying. They are two people describing two different jobs after forming two different opinions about the same door.
Get both options from one visit instead. One assessment, both figures, in writing. Then the comparison is real, and you can take that pair to somebody else if you want a second opinion on both.
“What would you expect to fail next on this door, and roughly when?”
Somebody who has assessed the whole assembly answers immediately, because they looked at all four clocks. Somebody who looked only at the broken part cannot, and you will hear it in the vagueness. The answer to that question is essentially the repair-or-replace decision.
Usually not — repairing it to sound and safe is the more common answer, because cost per remaining year is a poor guide when the remaining years belong to the buyer.
The honest caveat is that on most tract elevations here the garage door is the largest single surface visible from the street, so a tired door works against the house in photographs and viewings. Whether that justifies replacement is a question for your agent; whether the door is structurally sound is one we can answer for nothing.
Further reading
The diagnosis that comes first: which of the four assemblies failed, and whether the door is structurally finished at all.
The diagnosis guideIf the answer is replace: what to specify for your climate zone so the next one lasts.
The specification guideWhy we measure before quoting, what moves the figure, and what a written quote commits us to.
Our pricing policyThe practical offer
Where both options are genuinely viable, you get both in writing from the same visit: what the repair covers and explicitly what it does not, and what a replacement would involve. Plus the answer to the question above — what we would expect to fail next, and roughly when.
We will also say which we would choose and why, and that recommendation is frequently the repair. Six of the nine failures we are called out to most are part-level, so a door that needs one thing doing and has years left in the rest is the common case rather than the exception.
The measure and the assessment are free either way, and free whether you go ahead with either. We cover the whole of Riverside County from one shop in Riverside, and the person who answers the phone is the person who comes and looks at it.
Tell us what has failed and roughly how old the door is. If both options are viable you will get both quoted from the same visit, along with what we would expect to need doing next — which is usually the thing that settles it.
Riverside Classic Garage Doors · 6410 Industrial Ave #3, Riverside, CA 92504 · (951) 232-9778 · Monday–Saturday, 8am–6pm · Closed Sunday
Worth telling us when you ring