A coated garage floor either lasts fifteen years or starts lifting in two. The difference almost never comes down to the color you picked. Here's what actually decides it, from someone who gets called out to fix the ones that failed.
The call usually sounds the same. Somebody had their garage done a while back, it looked great in the photos, and now there's a patch by the door lifting like a sunburn, or the spot where the car sits has gone cloudy and started to flake. They want to know if it can be patched.
Usually it can't — not the way they're hoping. But the more useful question is why it happened, because the same three or four causes come up over and over, and every one of them is visible before you hire anybody.
Nearly everyone searching for this calls it an epoxy garage floor. I do too, on the phone, because that's the words people use. It's worth knowing what's actually going down on the slab, though, because it's part of the answer.
Epoxy is a specific two-part resin. It was the standard for years, and a professionally installed epoxy floor is a real upgrade over bare concrete. But it cures hard and stays hard. In an Ohio garage that's a problem: hard coatings go brittle when the slab is cold, and a car that's been running for an hour parks a set of hot, soft tires on top of them. That combination is what installers call hot-tire pickup — the tire grabs the coating and lifts a patch off with it when you pull out.
What we install now, and what most serious installers around Cleveland have moved to, is polyaspartic. It's in the polyurea family, it cures in hours instead of days, and it stays flexible in the cold, so it moves with the slab through freeze-thaw instead of cracking off it. Same flake look, same idea, very different behavior in February. So when I talk about coatings failing below, understand that some of that failure is baked into the material choice itself.
This is the big one. If I had to bet on why any given failed floor failed, I'd bet here first.
A coating doesn't glue itself to concrete by chemistry alone. It needs a mechanical key — a rough, open surface it can lock into. You get that by running a diamond grinder over the whole slab until the smooth troweled skin is gone and you're into clean, open concrete with a texture roughly like 60-grit sandpaper. It's dusty, slow, unglamorous work, and it's most of the labor in the job.
It's also the easiest step to fake. The shortcut is acid etching — mop on an acid wash, rinse, call it prepped. It costs almost nothing and takes twenty minutes. On a smooth garage slab it does not open the concrete anywhere near enough, and on a slab that's had oil dripping on it for fifteen years, it does nothing at all where it matters.
How to spot it after the fact: find a piece that lifted and turn it over. If the underside is smooth and glossy — like the back of a sticker, like a mirror — that coating was never mechanically bonded to anything. It sat on top of the slab, and it left the slab as clean as the day it went down. If instead the back is covered in gray concrete crumbs, the bond actually held and it was the top layer of the concrete itself that gave out. That's a different, rarer problem, and it usually means weak or badly finished concrete.
How to spot it before you hire: ask what the prep is, and listen for the word grinding. Then ask what happens to the dust. A crew that grinds properly shows up with a machine hooked to a serious vacuum. A crew planning to acid etch shows up with jugs.
The other half of the failures I see are material. Big-box DIY kits are the obvious version — a thin water-based epoxy, a bag of flakes, one afternoon of work. They look good for a season. They come up in sheets under a hot tire, and they come up early in the corner by the door where the salt water pools.
But it's not only the kits. Plenty of contractors are still putting down straight epoxy systems on residential garages because the material's cheaper and they've always done it that way. In a heated basement or a warehouse that stays at 65 degrees, fine. In a garage that swings from 90 in July to below freezing in January, with brine dripping off the car all winter, you're asking a rigid coating to do something it isn't built to do.
One clarification worth making, since it comes up on every estimate: road salt doesn't dissolve a good coating. Salt and brine can't eat polyaspartic. What they do is find whatever weakness already exists — a pinhole, a thin edge, a spot the grinder missed near the wall — sit in it through a hundred freeze-thaw cycles, and pry it open. Salt doesn't cause the failure. It just finds it and makes it visible by March.
This is the one almost nobody checks, including some people who do this for a living.
Your garage slab sits on the ground. If there's no vapor barrier under it — and on a lot of older slabs around here there isn't — ground moisture moves up through the concrete as vapor, all day, every day. You can't see it. The slab looks bone dry. Put a coating over it and that vapor now has nowhere to go, so it collects underneath and pushes. You get blisters that look like bubbles under wallpaper, or a floor that peels for no visible reason two summers later.
There are actual numbers for this, not opinions. The industry limits are a moisture vapor transmission rate no higher than three pounds per 1,000 square feet per 24 hours, tested per ASTM F1869, and internal slab relative humidity no higher than 75%, tested per ASTM F2170. Above that, you need a moisture vapor barrier primer before anything else goes down, or the floor is on a countdown from day one.
I check every slab with a meter before I quote it, and I don't check it in the middle where it's easy. I check the low spots — around the floor drain, in the corners, along the wall where the grade outside is higher, anywhere water would sit if you spilled a bucket. That's where the reading turns up if it's going to. Seasonal matters too: a slab that tests fine in a dry August can read very differently in a wet spring, which is why the barrier goes in when there's any doubt.
This one is seasonal and it's the reason some fall installs fail.
Polyaspartic has a real working window, and the manufacturer publishes it: air and surface both between 35°F and 85°F, with the material itself kept between 60°F and 85°F. Below 35 on the slab, you don't coat, period — no heater, no exceptions, no "it'll be fine." And below 60°F you have to switch to a faster-curing catalyst, because the slow one simply won't kick properly in the cold.
That last detail is where cold-weather jobs quietly go wrong. A crew that runs one catalyst year-round and doesn't adjust for a 40-degree garage in November can produce a floor that looks finished, feels finished, and never fully cured. It stays slightly soft, marks easily, and starts letting go the following summer.
If you already have a coated floor and want to know what you're dealing with, or you have bare concrete and want to know what a contractor will be facing:
Yes, but not by coating over it. Whatever is already down there becomes the weakest layer of the new floor, so if it's peeling now, it will take the new coating with it later. The old system has to come off — ground back to clean concrete across the whole slab, not just the failed patches, because a partial removal leaves you with a floor that's bonded in some places and not others.
Removal is real work and it's reflected in the price. It's still cheaper than doing it twice. Most of the redo jobs I take on are floors somebody paid for two or three years ago, and the honest math is that the original cheap quote ended up costing more than the right one would have.
Not a sales pitch so much as the checklist, since it's the same one you'd use to judge anybody:
We work out of North Royalton and cover the southwest Cleveland suburbs — Parma and the towns around it. If you want to know what your slab is actually dealing with, that's what the free estimate is for. See what a floor costs if you want numbers first, or how the two coatings compare if you're still choosing a system.
We'll test it, tell you straight, and quote it fixed — free, no pressure.
Get Your Free QuoteIn almost every failed floor we're called out to, the concrete was never mechanically ground before coating, or the coating was a hardware-store epoxy that can't handle hot tires and freeze-thaw. Slab moisture is the third cause, and it's the one nobody checks.
No. Anything left on the slab becomes the weak layer in the new floor. The old coating has to be ground off back to clean concrete first — across the whole slab, not just the failed patches.
Turn over a piece that lifted. Smooth and glossy underneath means the concrete was never opened up. Gray concrete crumbs stuck to the back means the bond held and the top layer of the slab is what gave out.
Salt doesn't dissolve a properly installed coating. It finds existing weak spots — pinholes, thin edges, a place the grinder missed — and freeze-thaw widens them. Salt exposes bad prep rather than causing failure on its own.
It goes by the slab, not the forecast. The material's working window is 35°F to 85°F on both air and surface, and below 60°F it has to run on a faster-curing catalyst. In practice, an unheated Cleveland garage sits below that window for most of the winter, so we don't coat residential garages in freezing weather at all — heated commercial space is a different story.