On a Tile Roof, the Tile Is Not What Keeps the Water Out
Tile sheds and wears. The underlayment underneath it waterproofs the roof, and it usually fails first. What Hill Country homeowners should specify.
A tile roof that is leaking with every tile intact is not a contradiction. It is the normal way a tile roof fails. Clay and concrete tile is a shedding layer and a wear surface, not the waterproofing. The membrane underneath the tile is what actually keeps water out of the house, and on most tile roofs that membrane wears out long before the tile does. The tile gets blamed because it is what you can see. The part doing the real work is the part nobody looks at until the ceiling stains.
This is not a defect in tile as a material. It is how the assembly has always worked, going back to the clay roofs it was copied from. Wind-driven rain gets past individual tiles constantly, especially at the overlaps, and the roof is designed around that. Water that passes the tile runs down the underlayment to the eave and off the roof, the same way water runs down housewrap behind siding. The tile’s job is to break the force of the rain and take the sun, the foot traffic, and the weather so the membrane underneath does not have to. When people say a tile roof typically lasts fifty years or more, they usually mean the tile does. The roof as a functioning waterproof system is only as durable as whatever is under it, and that is the number that actually decides when you write a check.
Why the tile outlives the roof underneath it
Fired clay and cured concrete tile are close to inert. Barring a cracked piece from foot traffic or a large hailstone, the tile itself does not really wear out on any timeline a homeowner plans around. A basic organic-felt or standard fiberglass underlayment does not come close. Under a Hill Country deck baking through a Texas summer, a 15-pound or 30-pound felt typically turns brittle and starts cracking well inside fifteen years, sometimes sooner on a west-facing slope with poor attic ventilation. The tile above it looks the same the whole time, so there is nothing to see from the ground or even from a casual walk of the roof. The failure shows up as a water stain, a soft spot in the ceiling drywall, or a musty smell in a closet under the slope, usually years after the underlayment actually gave out.
This is why a tile roof gets torn off and relaid partway through its service life. The tiles come up, the underlayment is replaced, and if they are in good condition the same tiles go back down. The job exists because of the membrane, not the tile. Homeowners who budget for tile as a one-time cost are budgeting for the wrong event. The real recurring cost on this roof type is the reroof of the underlayment underneath a wear surface that is often still fine.
The underlayment options, and why heat is the deciding variable here
Three broad categories cover most tile roofs, and they differ mainly in how they hold up to sustained heat and how much abuse they can take during installation and foot traffic.
Asphalt-saturated felt, in 30-pound weight, is the traditional baseline. It is inexpensive and it works, but it has real limits under a tile roof in this climate. Deck temperatures under dark tile in full Hill Country sun regularly run well above ambient air temperature, and sustained heat is what drives felt toward brittleness. Felt also has modest tear strength, which matters because a tile roof gets walked during installation, during service calls, and during its eventual relay, and every one of those trips is a chance to put a boot through a weak membrane.
Modified bitumen, specifically SBS-modified peel-and-stick membranes, are the upgrade most manufacturers now specify for tile in hot climates, and for good reason. SBS rubber keeps the membrane flexible across a wider temperature range than plain asphalt, the self-adhered bond means no fastener penetrations through the field of the membrane, and the tear and puncture resistance is substantially higher than felt. That last point matters as much as the heat tolerance, because a membrane that can shrug off a dropped batten or a careless boot is a membrane that survives the installation that puts it in place.
Synthetic underlayments are lighter, often more UL-rated for high temperature exposure, and generally offer good tear strength for their weight. Where they can fall short under tile specifically is at the details: many synthetics are built and tested as field membranes rather than as self-adhered products, so they still need a separate self-adhered membrane at the vulnerable spots rather than covering everything on their own. A synthetic underlayment is a legitimate choice on a tile roof, but ask what it is doing at the valleys and penetrations, because “synthetic underlayment” on a proposal does not by itself answer that.
None of these is universally wrong. A budget reroof on a house the owner plans to sell in a few years is a defensible place for a well-installed felt system, and saying otherwise would be dishonest. What is not defensible is not knowing which one is being installed, because the difference in service life between the categories is measured in decades, not years, and it is invisible once the tile goes back down.
Self-adhered membrane where the roof actually leaks
Most tile roof leaks do not happen in the open field of a slope. They happen at valleys, wall and chimney flashings, skylights, plumbing vent boots, and anywhere a low-slope transition slows water down instead of shedding it quickly. These are the spots where a self-adhered SBS membrane earns its cost even on a roof that uses a lighter-weight or synthetic field underlayment everywhere else. A hybrid approach, self-adhered membrane at the vulnerable details and a synthetic or felt underlayment in the field, is common and reasonable. What is not reasonable is a single layer of basic felt run straight through a valley with no additional membrane, which is a shortcut that shows up as a leak at exactly the spot you would expect.
Battens, the air gap, and what bad flashing does to it
Many tile assemblies, particularly higher-end and higher-profile tile, are installed over battens and counter-battens rather than laid directly on the underlayment. The counter-battens run with the slope and create a drainage channel and ventilation gap; the battens run across the slope and hold the tile. Done well, this air gap does two useful things: it lets any water that gets past the tile drain freely to the eave instead of sitting against the membrane, and it lets heat and moisture vapor move instead of being trapped against the deck. On a Hill Country roof, that ventilation is part of what keeps deck and underlayment temperatures from running even hotter than they already do under the tile.
The same system creates a real failure mode if it is flashed carelessly. Battens interrupt the drainage plane by design, and every place a batten crosses a valley, meets a wall, or approaches a penetration needs to be flashed so water is routed around the wood rather than trapped behind it. A batten set directly across a valley without a cricket or diverter, or nailed through the membrane without sealing the penetration, turns a ventilation feature into a dam and a row of holes. This is a detail worth asking about directly, because it is invisible once the tile is on and it is one of the more common causes of a leak that shows up at a batten line rather than at an obvious flashing point.
Fastening, wind, and where a tile roof actually comes apart
Tile fastening methods vary by profile and manufacturer, from nailed to wire-tied to foam-set systems, but the load is never distributed evenly across the roof. Wind uplift concentrates hardest at the perimeter, the eaves, the rakes, and the hips and ridges, the same way it does on any roof type. That is where fastening specifications tighten up, and it is where an installer under time pressure is most tempted to use the field spacing everywhere. A roof that looks fully fastened from the ground can still be under-secured exactly where the next real windstorm will test it first.
Flashing at walls, chimneys, and valleys: the other real leak source
Underlayment quality and batten flashing cover one category of failure. Sheet metal flashing at walls, chimney saddles, and valleys is the other, and the two interact. A tile roof with excellent underlayment and poor step flashing at a chimney will still leak at the chimney, because the flashing is what’s actually shedding water at that specific junction; the field underlayment is the backup, not the primary defense there. Copper is common at these details on higher-end Hill Country tile and slate roofs specifically because it lasts as long as the tile itself and ages well against masonry and stone. Whatever metal is used, ask how valleys are formed (open metal valley versus closed) and how step flashing is integrated with the underlayment at walls, because these are the joints where two different trades’ work has to agree with each other, and disagreement is where leaks start.
Walk it as little as possible
Tile breaks underfoot more easily than most people expect, and a cracked tile from a careless footstep looks identical to one cracked by hail. Every trip across a tile roof, whether it is the original installation, a satellite dish install, a gutter cleaning, or a home inspection, is a chance to crack tile that was fine the day before. A roofer who knows tile walks at the base of the tile near the headlap, distributes weight, and uses the fewest passes possible. One who does not leaves a trail of hairline cracks that will not show up as leaks for a season or two, at which point nobody remembers who was up there.
Lift-and-relay versus replacement: how to tell if the tile is worth keeping
When a tile roof’s underlayment reaches the end of its life, the tile itself often still has real service life left, and the decision is whether to lift and relay the existing tile over new underlayment or replace the tile along with it. The honest answer depends on the tile’s actual condition, not its age. Sound tile with no widespread cracking, no significant surface erosion on concrete tile, and a profile that is still available or has enough salvageable spares for breakage during removal is generally worth relaying; the underlayment beneath it is what was failing, not the tile. Tile with a high existing breakage rate, a discontinued profile with no reserve stock for the pieces that will not survive removal, or visible degradation across a large share of the field is a different conversation, because relay breakage on already-compromised tile can run high enough that replacing outright ends up being the more honest recommendation. Either way, this is a judgment call that should be made tile by tile, not decided from the ground.
What to ask for on a proposal
The membrane is the part of a tile reroof that determines how long the new work actually lasts, and it is also the easiest part to leave vague on a proposal. Before signing anything, get the underlayment specified by name and type, not just “underlayment”: felt weight if felt, or the specific SBS or synthetic product if not. Ask what is used at valleys and penetrations, and confirm it is a self-adhered membrane if the field underlayment is not. Ask whether the roof uses battens or counter-battens, and if so, how they are flashed at valleys and walls. Ask how the salvaged tile was evaluated, if any of the existing tile is being relaid, and what the plan is for breakage during removal. A proposal that answers those questions in writing is telling you what you are actually buying. One that just says “new underlayment installed” is not.
The Honest Roofer works on slate, clay, concrete and designer tile roofs across West Austin, West Lake Hills, Lakeway, Bee Cave, Spicewood and Horseshoe Bay, including slate and tile reroofs where the tile is sound and the membrane underneath it is not. The roofing materials overview sets tile alongside the other systems available, and the process page covers how a property gets reviewed before any material or membrane decision is made. If a tile roof on your property is leaking with nothing visibly wrong on top, what’s underneath is the place to look, and it turns on the same questions covered in the metal-versus-tile comparison and in what a real inspection should check after hail: what you cannot see is usually what decides the outcome.