Why did my pool tile fall off if the pool is not that old?

Age is often not the variable. The tile band sits at the junction between the pool shell and the deck, and the published pool tile assembly treats that line as a movement joint, sealed with a flexible material over a backer rod rather than grouted. If that joint is absent or was filled with grout, movement between the two structures has to pass through the tile, and the bond behind it is the weakest thing in the path. A young pool with a rigid joint can lose tile sooner than an old pool detailed correctly.

Can I just glue the loose tiles back on?

You can, and it is worth doing when the loss is genuinely isolated: a few scattered pieces with no pattern, on a pool with sound coping and no history of the same thing. What re-bonding does not do is change why the tile came off. If pieces released in a continuous run, or the same section has gone before, the cause is still present and the replacements are being placed back into it. In that situation the money is better spent establishing whether the bond beam and the coping joint are sound.

Is grout meant to flex at all?

No. The tile education body states that grout joints do not offer any movement-absorbing qualities, which is why movement joints are required to be filled with a resilient material instead, with silicone, urethane and polysulfide named as the examples. This is a property of the material rather than a question of quality or application. A correctly mixed, correctly cured grout is still rigid, because that is what grout is for. Where movement has to be accommodated, something else has to be in the gap.

Does glass tile last longer than ceramic at the waterline?

Not by virtue of being glass, at least where scale is concerned. The National Plasterers Council notes that all materials in a water-contact environment are subject to mineral scale or metal stain precipitation and lists glass surfaces explicitly alongside concrete, brick, metal and tile. Glass has real advantages in appearance and in resisting the water itself. It does not exempt an installation from water chemistry, and it does not change the movement question at the coping line, which is what most commonly takes a band off.

How often should waterline tile need replacing in Naples?

There is no published service interval, and any number offered is an estimate rather than a standard. What can be said is what shortens it: a rigid joint where a movement joint belongs, a bond beam that is moving, aggressive or scaling water concentrated at the one band that is alternately wet and dry, and cleaning aggressive enough to attack the grout. A band that has been detailed and balanced properly is usually replaced because the owner wants a different look, not because it failed.

Should the coping be replaced at the same time as the tile?

Not automatically, but it should be assessed at the same time, because the two share the joint that most often causes the problem. If the coping is hollow, rocking, lifted or its joint has opened, then tiling up to it and leaving it as found puts new work against a known moving element. Doing both together also means the sealant joint is formed once, properly, with a backer rod, rather than being disturbed again a season later. Where the coping is sound, tile alone is a reasonable scope.

Insights · Finishes, tile and stone

Why does waterline tile come off, and does re-tiling actually fix it?

Tile that has come off a waterline is usually described as having failed to stick, and the conversation goes straight to adhesive. The tile industry details the same location as a movement joint, sealed with a flexible material rather than grouted. That reframes the problem: the tile was not holding on badly, it was holding a joint shut.

Clayton Waggoner, Owner and Managing Member · Published August 8, 2026 · 12 min read

The waterline band is the strip of tile at the top of a pool shell, running just below the coping. It sits at the junction between two things that move differently, the shell and the deck, and it lives in the one part of the pool that is alternately wet and dry all day. Both of those facts matter more to how long it stays on than the mortar behind it.

The short version

I.The tile industry assembly for pools puts a flexible sealant over a closed-cell backer rod where the tile meets the coping. That line is designed to move.

II.Grout joints offer no movement-absorbing qualities, so grout cannot do a movement joint's job no matter how well it is mixed or applied.

III.Exterior tile assemblies are given movement joints at roughly 8 to 12 foot intervals in each direction, and more often where conditions demand it.

IV.The waterline is also the most chemically loaded band in the pool, because scale and etching both concentrate where water meets air.

V.Re-tiling restores the tile. If the joint detail and the cause of movement are not addressed with it, the new tile inherits the old problem.

Why does waterline tile come off in the first place?

Because something moved and the tile was the most brittle thing in the path. The tile band sits at the junction between the pool shell and the deck, two structures that expand, contract and settle independently. A rigid band spanning that junction takes the load.

Start with where the tile is. It is not in the middle of a wall. It is at the top of the shell, immediately under the coping, which is the line where the pool structure meets the deck structure.

Those two things are not one object. The shell is a bowl in the ground holding many tons of water. The deck is a surface on fill that heats in the sun, cools at night, and takes traffic. They move on different schedules and by different amounts.

The tile industry treats that line accordingly. In the pool assembly published for concrete, shotcrete and CMU tanks, the detail where the tile meets the coping and decking carries a flexible sealant over a closed-cell foam backer rod, not grout (TCNA method P602, 2020)¹.

A backer rod exists for one reason. It sets the depth and shape of a sealant bead so the sealant can stretch and compress without tearing. You do not put one behind something that is meant to stay still.

So the assembly anticipates movement at exactly the elevation where owners find their tile on the pool floor. The tile is not being asked to resist that movement. It is supposed to be sitting beside a joint that absorbs it.

When that joint is missing, or has been filled with grout, or has been sealed once and never maintained, the movement has to go somewhere. It goes into the most brittle element available, and the bond behind the tile is what gives.

Sunlight moving across the surface of pool water close to the waterline — The one band that is wet, then dry, then wet again, every day of its life.

Is grout supposed to hold the tile band still?

No, and this is the point most often missed. Grout has no capacity to absorb movement. The tile education body states it directly and specifies that movement joints be filled with a resilient material instead, naming silicone, urethane and polysulfide.

The sentence is unambiguous. Because grout joints do not offer any movement-absorbing qualities, movement joints must be filled with a resilient material, with silicone, urethane and polysulfide named as the examples (Ceramic Tile Education Foundation, in TileLetter)².

That is a material property, not a workmanship issue. A grout can be mixed correctly, cured correctly and still be the wrong thing in that gap, because it is rigid by design and a movement joint needs something that is not.

The spacing tells you how seriously the industry takes it. Exterior tile assemblies are given movement joints at 8 to 12 feet in each direction, with more frequent placement where materials and environmental conditions require it (CTEF, in TileLetter)².

A pool waterline is an exterior assembly in continuous water contact, in direct Southwest Florida sun, wrapping a structure that is itself under load. It is not the easy end of that range.

Responsibility is assigned too. The published method states that the design professional or engineer shall show the specific locations and details of movement joints on project drawings (EJ171, cited by CTEF)². On a residential remodel there is frequently no such drawing and no such professional.

That gap is worth naming plainly, because it explains why this failure is so common on otherwise competent work. The requirement is real, the drawing that should carry it usually does not exist, and the tile setter inherits a decision nobody made.

What does a correct pool tile assembly actually include?

More layers than most owners expect, and each one has a named performance standard. The published assembly runs waterproofing, a bonded mortar bed, a scratch coat, a bond coat and grout, with sealant at the movement joints and specified minimum standards for each material.

The assembly for a concrete, shotcrete or CMU pool tank is built up rather than stuck on. It carries cementitious waterproofing on the tank, a mortar bed bond coat, a mortar bed, a scratch coat, an optional waterproof membrane, a bond coat and then grout (TCNA P602, 2020)¹.

Each of those has a floor it must meet. The cementitious bond coat is specified at ANSI A118.4 or better, cementitious grout at ANSI A118.6 or better, and epoxy alternatives at ANSI A118.3 (TCNA P602 minimum performance requirements, 2020)¹.

Movement joints and sealant carry their own references, to EJ171 and ASTM C1193 for the joints and ANSI A118.01 with ASTM C920 for the flexible sealant (TCNA P602, 2020)¹.

Waterproofing membranes are ANSI A118.10 and crack isolation membranes are ANSI A118.12, and the cementitious waterproofing is specified as suitable for positive and negative hydrostatic pressure and for swimming pools (TCNA P602, 2020)¹.

That last phrase deserves attention on a Naples job. Negative hydrostatic pressure means water pushing from behind the surface rather than from the pool side, which is precisely what a high water table produces.

The point of listing all this is not to make a re-tile sound daunting. It is to show that the specification exists in public, with numbers, and that a proposal can be checked against it. A quote that names its bond coat standard is telling you something real.

When a shell is opened up far enough to touch the bond beam, this stops being only a tile question, which is where pool crack and shell repair and the tile scope meet on the same job.

What it is for

Why is the waterline the harshest band in the pool?

Because it is the only part of the surface that is alternately wet and dry all day. Everything dissolved in the water concentrates there as water evaporates off it, which is why scale and staining show at that line first and most heavily.

Scale is deposited from water onto whatever is present. All materials in a water-contact environment are subject to mineral scale or metal stain precipitation, including concrete, cementitious coatings, brick, metal, tile and even glass surfaces (NPC Technical Bulletin 4, 2018)⁴.

Glass tile is on that list, which is worth remembering when a mosaic is chosen partly on the promise that it stays clean. Scale does not care what it is landing on.

The chemistry cuts both ways at the same line. Low calcium hardness, low carbonate alkalinity or low pH create an aggressive environment that etches the cement in the areas of greatest solubility and porosity (NPC Technical Bulletin 3, 2016)³.

Those areas include the interfaces. Etched areas appear around aggregate particles, tile inserts and other interfaces, and often show as small spots (NPC, 2016)³. A tile edge is an interface, and grout between tiles is cementitious.

Salt systems raise the stakes slightly. With higher salt levels, the water can be more aggressive (NPC, 2016)³. That is not an argument against salt chlorination, which many Naples pools run happily, but it is a reason the water balance around a tile band is not a formality.

Put the two things together and the waterline is doing two jobs at once. It is the most chemically loaded strip of surface in the pool, and it is sitting on the one line in the structure that is supposed to move.

Does re-tiling fix the problem, or does it reset the clock?

It depends entirely on whether the cause was addressed with it. New tile on a restored joint detail and a sound bond beam is a durable repair. New tile bonded across the same rigid line, over the same movement, is the same assembly with a fresh surface.

Ask what is being replaced. If the scope is tile, thinset and grout, the work is cosmetic in the literal sense: it renews the visible layer and leaves the assembly beneath it as found.

That is sometimes correct. Tile does come off for ordinary reasons, including an old installation reaching the end of a genuinely long life, or a localized area disturbed by something specific.

It is not correct when the tile came off in a line, or came off twice, or came off along the same run each time. A pattern is evidence of a cause that is still present.

The questions that separate the two are practical. Was there a soft joint at the coping, and what was in it. Did the tile debond with mortar still stuck to the shell, or did it come away clean. Is the bond beam sound where the tile was.

Tile that releases with its mortar attached to the back and nothing on the shell points at the bond to the substrate. Tile that shears off leaving mortar behind points more toward movement or force through the band.

Where the answer is movement, the repair includes the joint. That means a real sealant over a backer rod at the coping line rather than grout, which is what the published assembly asked for in the first place (TCNA P602, 2020)¹.

Scope written this way costs more on the day and less over the life of the pool, and it is the same reasoning that separates a finish from its substrate in choosing a pool interior finish. The visible layer is rarely the thing that decided the outcome.

Water test vials resting on pool coping in afternoon light — Balance decides whether the tile line scales, etches, or is left alone.

How can I tell whether it is the tile or the structure underneath?

By looking at how the tile left, where it left, and whether anything else on the pool agrees with it. Tile is a surface, so it reports on what is behind it. A single loose piece and a whole run releasing along the coping are different findings.

Look at the back of a piece that has come off, if you still have one. Mortar on the tile with a clean shell says the bond to the substrate failed. A clean tile with mortar still on the shell says the failure was in the bond coat or through the tile itself.

Look at the line. Loose tile scattered without pattern reads as ordinary aging. A continuous run letting go along the top, especially on one side of the pool, reads as movement or a bond beam problem.

Look at the coping above it. Hollow sounding coping, lifted or rocking stones, or a joint that has opened all point at the same junction the tile is sitting on.

Look for company. Cracks in the shell at the same elevation, a deck that has settled or heaved nearby, or a spa wall doing something similar all raise the likelihood that the tile is a symptom.

Look at the history. Tile replaced once and gone again in a few seasons is close to conclusive: the cause outlived the repair, and repeating the repair will repeat the result.

None of that requires equipment. It requires being willing to conclude that the cheap answer is the wrong one, which is the harder part of this trade, and the reason waterline tile replacement is quoted here after the bond beam is looked at rather than before.

Because the tile band sits between the pool and the deck, it is frequently the layer that decides how wide a project has to be, which is where resurfacing against remodeling starts.

“The tile was not holding on badly. It was holding a joint shut.”

Interactive · Before you re-tile

Is this quote fixing the tile or fixing the cause?

Four questions to put to whoever is quoting. They are not trick questions, and a contractor who works to the published method will have ready answers to all of them.

Ask Clayton

Questions this article usually prompts

Tap a question. Clayton replies himself, here and on the phone.

Clayton Waggoner Owner and Managing Member · Splash Pools Online

Tap a question to ask

Not covered here? Call 239-450-5933, it reaches Clayton, not a call center

If this is the situation you are in

Waterline tile replacement, starting behind the tile →Structural crack repair, starting with what is causing it →Pool coping replacement, and the beam underneath it →Pool remodeling in Naples, handled as one project →

References

All articles →

The article is general.
Your pool is specific.

Tell us what the pool is doing, a sentence or two is enough. The person who wrote this is the person who calls back.

Call 239-450-5933 Book a consultation