How do you fill cracks in concrete?

By Dan Reyes, Concrete finisher, 18 years Last reviewed

Filling a crack starts with one question: does the crack move? Moving cracks need a flexible sealant that can stretch through seasonal opening and closing. Static cracks can take a rigid material — epoxy, polyurethane resin, or a patching mortar. Choosing the wrong one is why most crack repairs fail within a year.

Routed channel
¼ × ¼ intypical for sealant
Backer rod
over ½ in deeptwo-sided bond
Sealant standard
ASTM C920elastomeric joint sealants
Epoxy standard
ASTM C881epoxy resin bonding systems

Short version

  • Moving crack gets a flexible sealant; static crack gets a rigid filler.
  • Route the crack into a clean channel before filling anything. Prep decides the outcome.
  • Backer rod keeps the sealant bonded on two sides so it can stretch instead of tearing.
  • Epoxy injection restores structural continuity but only on cracks that are done moving.
  • A filled crack is a repaired symptom. Fix the cause or it comes back.

How do you tell whether a crack is moving?

A moving crack changes width with temperature, moisture, or load, and it will destroy any rigid filler you put in it. The simplest test is time: mark the crack width with a pencil line at each edge, date it, and check again in a different season. A crack that measures the same in January and July is static.

Two shortcuts help when you cannot wait six months. Any crack that crosses a control joint, an isolation joint, or the full width of an exterior slab is almost certainly a moving crack, because that is exactly what those slabs do — they expand in heat and contract in cold. Any crack inside a heated, conditioned interior slab that has been in place for years and shows no offset is probably static.

Crack monitors, small two-piece gauges that mount across a crack and read displacement directly, cost very little and remove the guesswork on a crack that matters.

Cracks that are still actively widening — as opposed to opening and closing seasonally — are a different problem entirely. Something is still moving, and no filler is the answer until it stops.

What materials fill a crack in concrete?

Five material families cover nearly every job, and the choice comes down to movement, width, and whether the repair has to carry load.

MaterialUse it forMovement
Elastomeric sealant (ASTM C920)Slab and joint cracks, exterior flatworkHandles movement
Pavement joint sealantDriveways, sidewalks, wide exterior cracksHandles movement
Epoxy injection (ASTM C881)Structural cracks in walls, beams, columnsStatic only
Polyurethane injectionLeaking cracks in foundation wallsSome movement
Patching mortar or polymer-modified repair mortarSpalled edges, wide static cracks, cornersStatic only

Sealants and pavement sealants stay flexible for the life of the repair and are the default for anything outdoors. Epoxy is the opposite — cured epoxy is stronger than the surrounding concrete, so an epoxy-injected crack will simply crack again beside the repair if the member keeps moving. Polyurethane injection foams on contact with water, which is why it is the standard answer for a wet, leaking basement wall crack rather than a structural one.

Patching mortars fill volume. They rebuild a broken edge or a spalled corner, but they are not sealants and they will not tolerate movement.

Why does surface preparation decide the result?

Preparation is where crack repairs are won and lost, and it is the step everyone shortcuts. Any filler bonds to the surface it touches, so if that surface is dust, laitance, curing compound, old sealant, algae, or damp concrete, the bond is to a weak layer that peels away with the repair still stuck to it.

Rout the crack first. A crack chaser wheel on an angle grinder, or a masonry blade run along the crack, opens a channel roughly ¼ inch wide and ¼ inch deep with parallel sides. Routing does two things: it removes the feathered edges where the concrete is weakest, and it gives the sealant a shape with enough cross-section to stretch.

Then clean aggressively. Vacuum, blow with oil-free compressed air, vacuum again. Wire-brush the channel walls if there is any old material. Let it dry unless the product specifically calls for a damp substrate.

Prime if the data sheet says to prime. Several high-performance sealants need a primer on concrete and will fail without it, and no amount of good technique substitutes for reading the product’s instructions.

Dry-cutting and grinding concrete release respirable crystalline silica, regulated by OSHA under 29 CFR 1926.1153.

When do you need backer rod?

Backer rod belongs in any channel deeper than about half an inch, and it does something specific: it forces the sealant to bond to the two side walls only. A sealant stuck to both sides and the bottom is bonded on three faces and cannot deform when the crack opens, so it tears internally or pulls away from one side.

Use closed-cell polyethylene rod sized about 25 percent larger than the channel width so it compresses into place and stays put. Press it to a uniform depth — most sealant manufacturers want a joint profile roughly twice as wide as it is deep, and the rod is how you control that depth.

Bond breaker tape does the same job where the crack is too shallow for rod. A strip of tape along the bottom of the channel keeps the sealant off the base.

Neither one is optional on a joint that moves. Three-sided adhesion is the most common cause of sealant failure in flatwork, and it is invisible until the sealant splits down the middle a year later.

What is epoxy injection and when is it the right call?

Epoxy injection restores structural continuity across a crack by forcing low-viscosity resin into it under pressure until the crack is full from face to face. Once cured, the member behaves as though the crack were not there — cured epoxy meeting ASTM C881 is stronger in tension than the concrete around it.

That strength is also the limitation. Epoxy injection only works on cracks that have stopped moving, because a repaired crack that keeps moving simply cracks somewhere new. Before injecting, the underlying cause has to be identified and corrected.

Injection is specialized work. Ports are set along the crack, the surface is capped with a paste, resin is injected from the lowest port upward until it appears at the next one, and the caps are ground off afterward. ACI and ICRI publish the practice guidance, and both treat it as a repair designed by someone who understands why the crack formed.

For a homeowner with a cracked garage slab or driveway, epoxy injection is almost never the answer. Flatwork on grade needs to move. Driveway-specific diagnosis and repair is covered on the concrete driveway crack repair page.

General information, not engineering advice. Structural work should be designed by a licensed engineer.

Does filling a crack fix the problem?

Filling a crack treats the symptom, and whether that is enough depends entirely on why the crack formed. Drying shrinkage cracks — the most common kind — are done moving in any meaningful sense once the concrete has dried, so sealing them keeps water and salt out and that is the whole job.

Cracks from an inadequate or missing control joint pattern will keep appearing until the slab has enough joints to relieve the shrinkage. NRMCA’s guidance on joints in slabs on grade covers spacing; the convention is 2 to 3 feet per inch of slab thickness, cut a quarter of the depth.

Cracks from settlement, frost heave, tree roots, poor drainage, or an undersized slab come back after any repair, because the force that opened them is still there. Correct the drainage, remove the root, lift the slab, or replace the panel.

The value of sealing even a symptomatic crack is real, though. Water entering a crack corrodes reinforcement, freezes and widens the crack, and washes out fines from the base underneath. Sealing buys years even when it does not solve anything.

Common mistakes

  • Putting rigid filler in a moving crack — it re-cracks along the repair, usually within one temperature cycle.
  • Skipping the routing step — feathered sealant on an unopened hairline has no cross-section and no bond area.
  • Omitting backer rod — three-sided adhesion tears the sealant from the inside and the failure is invisible until it splits.
  • Injecting epoxy into a crack that is still moving — the member cracks again beside the repair and the money is gone.
  • Treating the fill as the fix — settlement, heave, and missing joints reopen the crack no matter what went into it.

Also asked as

  • how do you fill cracks in concrete
  • how to fill concrete cracks
  • how to fill concrete crack
  • how to fill a concrete crack
  • how to fill in cracks in concrete
  • how to fill cracked concrete
  • how to fill in concrete cracks
  • how to fill a crack in concrete
  • how to fill crack in concrete
  • can you fill concrete cracks
Sources (5)
  1. NRMCACIP 4 — Cracking Concrete Surfaces (accessed Fri Aug 14 2026 00:00:00 GMT+0000 (Coordinated Universal Time))
  2. NRMCACIP 6 — Joints in Concrete Slabs on Grade (accessed Fri Aug 14 2026 00:00:00 GMT+0000 (Coordinated Universal Time))
  3. American Concrete InstituteACI 224R and 224.1R — control and causes of cracking in concrete (accessed Fri Aug 14 2026 00:00:00 GMT+0000 (Coordinated Universal Time))
  4. International Concrete Repair InstituteICRI — concrete repair guidelines and technical resources (accessed Fri Aug 14 2026 00:00:00 GMT+0000 (Coordinated Universal Time))
  5. ASTM InternationalC920 and C881 — joint sealants and epoxy resin bonding systems (accessed Fri Aug 14 2026 00:00:00 GMT+0000 (Coordinated Universal Time))

Related questions