Repair or Replace?
Repair when it's safe, when the fix restores real function, and when the cost stays well below replacement. Replace when safety is on the line, when you've already fixed it twice, or when repair costs more than half of new. Sometimes the right repair is letting go.
Every broken thing on your bench asks the same silent question: is this worth fixing? Most of us answer with our gut — or with whatever the nearest advertisement suggests. This guide offers five honest tests that sort the repairs worth doing from the ones that will only cost you twice.
It will not tell you to repair everything. A company that makes repair products has every incentive to say "fix it all" — we won't. Some things should be replaced, and saying so is part of honest repair culture. The goal is never regretting the choice.
Safety first: the red line is not negotiable
Before cost, before sentiment, before anything: ask whether the part keeps someone safe. Brakes, steering, load-bearing mounts, pressure vessels, fuel and brake lines — these are not places to economize. A failed repair here doesn't cost money. It costs far more.
If a safety-critical part breaks, there are only two honest answers: replace it, or hand it to a qualified professional. There is no third option where you save fifty dollars and hope. Hope is not a repair strategy.
The red line below is drawn conservatively on purpose. If you are unsure whether something is safety-critical, treat it as if it is. Caution here is free; the alternative is not.
The cost test: the honest math
Add up the true cost of the repair: parts, materials, and any tool you must buy for this job alone. Then compare it against a good replacement — not the cheapest knockoff, not the premium flagship, the honest middle.
The rule of thumb: if the repair costs less than half of replacement, the math favors repair. Above that, replacement usually wins — every repair carries a failure risk that a new part doesn't, and that risk has a price.
Count every attempt. A twenty-dollar fix that fails twice is not a twenty-dollar fix — it's sixty dollars and two weekends.
The function test: will it truly work again?
A repair is only a repair if the thing works properly afterward. A chair leg glued at a slight angle is not repaired — it's a wobble you built yourself. A pipe patched but still seeping is a leak with extra steps. Ask: after the fix, will this do its job at full strength, or will I be working around it forever?
Cosmetic fixes are fine when cosmetics are the job. But when function is the job, "good enough" is a loan you'll repay with interest.
Be especially honest about structural repairs. A bracket that holds "probably" is not holding. If you cannot test the repair under real load, you cannot call it done.
The time test: your hours have a price
Six hours of careful work to save fifteen dollars is not thrift — it's a hobby. And hobbies are wonderful, as long as you know which one you're doing.
Price your time honestly — not at your salary, but at what those hours are worth to you on a Saturday. If the repair takes longer than the thing is worth to you, replacement is the rational answer, and there is no shame in it.
Time spent learning is never wasted — your first epoxy repair teaches you the next ten. But a safety-critical part is the wrong classroom; practice on the chair leg, not the brake line.
The sentiment test: some things outrank math
Your grandfather's chair. The mug your kid made. The tool that's been in the family longer than you have. These things fail every cost test and deserve repair anyway — because their value was never in the price tag.
Sentiment is a legitimate reason, not a legitimate excuse. It doesn't override the safety test — a sentimental brake rotor is still a brake rotor — and it doesn't guarantee success.
When sentiment says repair and math says replace, do the repair properly. Use the good materials. The thing earned it.
The environmental ledger: repair, but honestly
Every repair keeps mass out of the landfill and avoids the energy of manufacturing something new. That matters — but a failed repair wastes twice: the materials you consumed attempting it, plus the replacement you buy anyway. The greenest repair is the one that holds; a rushed repair on the wrong substrate is litter with extra steps.
So the test is simple: repair the things you can repair well; replace — and recycle — the things you can't. Our glue and surface-prep guides exist to move more items into the first column.
The "twice" rule: two strikes and it's out
If you've repaired the same thing twice and it failed again, stop. Replace it. This isn't superstition — it's pattern recognition: the first failure might be bad luck, but the second is information. A third attempt is usually just the second failure with more hope attached.
The rule has one exception: if the second attempt failed for a reason you can now name precisely — wrong adhesive, skipped surface preparation, loaded it too early — then one corrected attempt is reasonable. But that's a new repair with a diagnosis, not the same hope repeated.
Letting go is also a decision
Replacement is not defeat. Sometimes it's the most honest form of repair: admitting that this object's working life is over.
Letting go well matters too: strip the usable parts, recycle the metal, keep the screws — and note what killed it, so the replacement lasts longer. The question was never really "repair or replace." It was always "what's the honest answer for this thing, this time?"
The five tests at a glance
| Test | The question | Repair when… | Replace when… |
|---|---|---|---|
| Safety | Could a failure hurt someone? | It couldn't — no safety role | It could — no debate, replace or call a pro |
| Cost | Does the repair cost less than half of new? | Yes, comfortably inside the repair zone | No — or the meter is already running on attempt two |
| Function | Will it work like new afterward? | Full function restored and testable | It will always be a workaround |
| Time | Are my hours well spent? | The fix is quick, or the work itself is joy | Hours of labor to save pocket change |
| Sentiment | Is it irreplaceable? | Its value was never in the price tag | Never — not over safety |
Six calls, made honestly
| The thing | The situation | The honest answer |
|---|---|---|
| Ceramic mug, clean break | Epoxy, ten minutes, full function restored | Repair |
| Brake caliper bracket, cracked | Safety-critical — failure means no brakes | Replace — or hand it to a pro |
| 12-year-old washer, third breakdown | Two repairs already failed | Replace — the twice rule |
| Grandfather's oak chair, split leg | Sentimental, and wood glues beautifully | Repair — properly, with good materials |
| Polypropylene storage bin, split | Adhesive won't bond PE/PP — see our glue guide | Replace — or use screws |
| Phone with a swollen battery | Fire risk — puncturing it is dangerous | Replace — do not glue or puncture the battery |
Frequently asked questions
Is it ever worth repairing something cheaper than the repair itself?
Yes — when sentiment or learning is the real product. Just be honest that you're paying for meaning, not savings, and give the repair the quality the meaning deserves.
What exactly counts as safety-critical?
Anything whose failure could injure someone or cause major damage: brakes, steering, structural mounts, pressure vessels, fuel and brake lines, electrical mains, stressed engine parts. When unsure, treat it as critical — the cost of being wrong in this direction is zero.
My appliance is ten years old. Repair or replace?
Run the tests in order. Check parts availability — a repair you can't get parts for isn't a repair. Apply the twice rule. And weigh running efficiency honestly: a modern replacement is often significantly more efficient, which belongs in the cost math.
The part is discontinued. Now what?
Discontinued parts make repair relatively more attractive — but don't spend more chasing a ghost than the thing is worth. Check donor units and aftermarket suppliers, or whether the part can be rebuilt. If the hunt costs more than a good replacement, let the hunt go too.
How do I put a price on my own time?
Not at your salary — at what those hours are worth to you. And separate learning from chore time: hours that teach you a reusable skill are an investment; hours redoing a failed fix are just cost.
I repaired it and it broke again. One more try?
Only with a diagnosis. If you can name exactly what went wrong — wrong adhesive, skipped prep, loaded too early — one corrected attempt is reasonable. Otherwise a third attempt is hope, not repair. Replace it.
Sources
- FAA Advisory Circular AC 43.13-1B, Acceptable Methods, Techniques, and Practices: Aircraft Inspection and Repair — the free public reference for bonded repairs: https://www.faa.gov/documentLibrary/media/Advisory_Circular/AC_43.13-1B_w-chg1.pdf
- West System, Instruction Manuals and Technical Guides — judgment from the boatyard world, where the question is asked daily and answered in fiberglass: https://www.westsystem.com/
- iFixit, the free repair manual for everything — the largest public archive of repair guides — living proof that repairability is a design choice: https://www.ifixit.com/