The honest answer is that most garage doors are worth repairing, and that a meaningful minority are obviously finished. Age on its own is a poor guide, because a twenty-year-old door with one history and a forty-year-old door with five are not comparable. What decides it is the state of the frame, the tracks, and the door itself.
When repairing is clearly the right call
- One part has failed and the rest of the system is sound. A cable, a spring, a roller, an opener fault, or one damaged section.
- The frame and tracks are square and sound. This is the single most important factor, because a good door in a good opening is worth a lot.
- The door runs true. It moves smoothly, the gaps are even, and it is not catching anywhere in the travel.
- Only the springs or cables need replacing and the springs can be correctly matched to the door.
- The repair cost is small relative to a replacement. Most of the common faults are a fraction of the cost of a new door.
A door in this condition is not “old and failing.” It is a good door with a part that has reached the end of its life, which is an ordinary thing and an ordinary repair.
When replacement is genuinely the answer
- Panels delaminating across several sections. The core is going; a new section will not match what is left.
- Multiple sections rusted through or rotted. The life is gone whatever you patch.
- Tracks rusted through, bent, or no longer square. This is what usually tips an old door over, and tracks are part of the door system rather than an optional extra.
- Springs replaced with the wrong size at some point, so the door is now permanently heavy and the springs keep failing.
- The frame or header has moved, so the door does not sit square and never will.
- It is a single-piece or warped slab door. These do not repair, and they are the least secure option in a residential garage.
- No insulation, and the garage is a space you actually use. A single-skin steel door is a large thermal break in a heated space.
- Repeated repairs, where you are now on the third visit for the same system in a few years.
The 50% rule — and why it is only a starting point
There is a rule of thumb in this trade that is genuinely useful, and it is worth knowing even if you never use it as a decision on its own:
The logic is sound. Once you are past the halfway mark, the cost of the individual repair starts to climb faster than the discount you would get on a new door, because you are also paying labour twice and you keep the old door’s problems. A door needing three sections is not cheaper to fix than one needing four.
It also has a sensible second use: it tells you when to stop negotiating repairs. If a technician is proposing a third section on a door that already has two bad, you are past the line, and it is fair to ask what replacement would cost in comparison rather than agreeing to another patch.
Why we do not rely on it alone
It counts components, not condition, and those are not the same thing. Two honest limits:
- Tracks are not counted. A door with two rusted-through sections and rusted-through tracks is well past 50% by any honest measure, but the rule as written would call it a repair. This is the most common way people get talked into a repair they should not have.
- Safety is not a percentage. A single badly fitted or mis-sized spring is a non-negotiable safety fault and gets replaced regardless of what else is wrong. “Only 20% of it is broken” does not mean 20% of the risk.
So use it as a checkpoint in the conversation, not as the answer. Ask what proportion of the door is being replaced and whether the tracks and frame are being assessed at all. If a quote does not mention the tracks, that is the question to ask before you sign anything.
What the decision actually depends on
| Factor | Points to repair | Points to replace |
|---|---|---|
| Frame and header | Square, plumb, sound | Racked, sagging, or rotted |
| Tracks | Square and surface rust only | Bent, pitted, or rusted through |
| Springs | Correctly rated, or replaceable | Previously mis-sized, door is heavy |
| Panels | Sound, or one section damaged | Delaminating across several |
| Failure history | One part, once | Same system, repeatedly |
| How the door runs | Smooth, square, quiet | Binds, rubs, or sits crooked |
| Insulation | Already insulated | Single skin, in a used space |
| Cost to the door | Repair is a fraction | Repair approaching replacement |
Count how many rows point at replacement. That is a much better guide than instinct, and it is the framework we use.
Two things worth being careful about
Two practical protections either way: ask for the work to be itemised, so you can see what is included; and ask what happens if the repair does not hold, because a person confident in the work will tell you.