How to Tell If a Hollow Tile Needs Replacing Instead of Just Injecting Resin

A hollow tile isn't automatically a bad tile. Injection repair works by pumping a low-viscosity epoxy or polyurethane resin through a small drilled hole into the air gap under the tile, re-bonding it to the substrate without lifting the tile itself. That only works if the gap is shallow and the tile is still structurally sound — which is why the size of the hollow area, how many tiles are involved, and whether the tile is cracked matter more than the hollow sound by itself.

The thresholds installers work from: hollow area covering 25% or more of a single tile's face, three or more adjacent tiles sounding hollow together, a crack wider than 1/16 in (0.0625 in) running through the tile, or a void deeper than 1/8 in (0.125 in). Cross any one of those and injection stops being a repair and starts being a way to postpone a replacement that's coming anyway.

Tile material changes the math. Natural stone with a hairline crack under the 0.0625 in threshold can often be injected and the crack sealed with a color-matched filler, because stone has some give. Porcelain with any through-crack near a shower curb or entryway threshold keeps propagating under foot traffic instead, since porcelain has almost no flex — a crack that would be fine to inject in stone is a replacement candidate in porcelain at the same width. Location matters just as much: if the hollow spot sits over a failed waterproofing membrane rather than just a thinset void, injecting resin traps moisture against that membrane and speeds up delamination in the tiles around it. Replacement is the only way to actually look at the membrane underneath.

Traffic and history matter too. In a high-traffic entryway or kitchen aisle, even a hollow spot under the 25% threshold tends to crack eventually because repeated point loads concentrate stress at the edges of the void, so installers often replace sooner than the numbers strictly require. If the grout around several tiles has already started cracking on its own, that's a sign the substrate itself is moving — injecting resin into each affected tile treats the symptom, not the cause, and the run of tiles usually needs to come out together rather than one at a time.

Resin can't reach every corner of a deep or irregular void, especially behind grout lines, so a tile injected past the 0.125 in depth limit keeps rocking on whatever contact points the resin did reach. The grout ring around that tile cracks again within months, and the hollow sound comes back — except now there's cured resin under the tile making removal messier than it would have been the first time.

In wet areas the failure is worse than cosmetic: trapped moisture behind a compromised membrane feeds mold growth that isn't visible until the tile is finally pulled. And financially, injecting a tile that actually needed replacing means paying twice — once for the injection, again for the eventual demo and reset. If the decision leans toward replacement, size the affected square footage in the /floor-tile-calculator before ordering material, and run the job through /guides/calculating-tile-replacement-costs to see labor and demo costs separately from the tile itself.

Injection makes sense for an isolated single tile under 25% hollow, with no crack and no wet-area membrane concern — particularly when the tile is discontinued or hard to match, since injection avoids opening up a matching problem entirely. The procedure itself is straightforward when the tile qualifies, and it's covered step by step in the how-to on fixing a hollow tile without removing it.

Either way, plan for the tile's remaining life rather than just the current spot. Run the room's total tile count through /tile-estimator so a few boxes of the same batch stay on hand as attic stock — a successful injection today doesn't guarantee the tile next to it won't fail in five years, and matching a discontinued run later is harder than keeping spares now.