HyreFoundation

Diagnosis

Which foundation cracks actually matter

Most cracks mean nothing. Width, change over time and whether the structure itself has moved tell you which ones do.

Updated August 16, 2026

Find a Foundation Professional What it costs

Written by HyreFoundation Research Desk Primary-source research and fact checking

~5 mm the threshold engineers flag Burland damage scale, categories 2 to 3 (Mair, Taylor & Burland, 1996, ISSMGE)
1% floor tilt that is usually noticeable ASCE Texas Section residential foundation guidelines, Version 3 (2022)
L/360 deflection criterion engineers use ASCE Texas Section residential foundation guidelines, Version 3 (2022)

The short answer

Most foundation cracks are harmless. The damage scale engineers publish grades cracks by width, not direction: under 1 mm is very slight, up to 5 mm is slight, and about 5 mm is the important threshold.

Its authors warn that width alone is not a measure of severity. A crack that is widening, or doors, floors and walls that have moved, matter more.

If this is happening now

A wall that has moved suddenly, a door frame visibly out of square, a floor that has dropped, or masonry that is leaning or bulging rather than just cracked — those are not monitoring situations.

That is the point at which a structural engineer should look at it before anything else happens. You can also reach HyreFoundation on (866) 582-8523.

Key takeaways

The published scale runs in six categories from hairline (under ~0.1 mm) to very severe (over 25 mm), and its authors say the important threshold is between categories 2 and 3 — around 5 mm
The same scale carries a printed caveat: "Crack width is only one factor … and should not be used on its own as a direct measure" of damage
The vertical / diagonal / horizontal rule repeated across the internet is not in any code or standard we could find

We looked, and we say so below.

Change over time is the most diagnostic thing available to you

, and it is free. Two dated photographs months apart beat one alarming photograph.

A floor out of level by more than about 1% is usually noticeable; engineers assess foundation deflection against an L/360 criterion, not against your cracks

What is actually diagnostic, in order

Roughly in order of how much it tells you, which is close to the reverse of the order it usually gets discussed in.

Whether it is changing

A crack the same width since 2019 is a historical event. One that opened 2 mm this summer is an active process. This is the most informative thing you can establish, and it costs nothing.

Whether the structure, not just the finish, has moved

Doors and windows out of square, floors sloping, walls leaning or bulging, separation at wall-to-ceiling junctions. Burland’s categories 4 and 5 are defined largely by these, not by crack width.

How many, and where

A pattern of cracking concentrated at one corner of a house says something a single crack does not.

Width, measured rather than estimated

A crack card or a set of feeler gauges costs very little. "About a quarter inch" is usually 4 mm and sometimes 9 mm, and those are different categories.

What changed outside

A tree removed or planted, a drought or an unusually wet year, a new patio or driveway, irrigation running against the wall, a plumbing leak, or excavation next door. Soil moves because water does; see why foundations move.

Whether the house is new

Concrete shrinks as it cures and most slabs develop shrinkage cracking in the first year or two. That is expected, not settlement.

How to monitor a crack properly, for almost nothing

Monitoring is the highest-value thing on this page. Evidence of change is what turns an anxious guess into an engineering question with an answer.

01
Photograph it with scale and a date

Put a ruler or a coin against the crack and photograph it square-on, not at an angle. Write the date on paper in the shot; photo metadata gets stripped and argued about. Do this for every crack, inside and out.

02
Measure the width at a marked point

Cracks are not a constant width. Mark a specific point with a pencil and always measure there. A crack comparator card costs a few dollars and removes the guesswork that makes the reading useless.

03
Mark the ends

A pencil tick at each end of the crack, dated. A crack that lengthens is telling you something that a crack that merely looks bad is not.

04
Repeat across a season change

Expansive clay cycles with moisture, so cracks that open in a dry late summer and close in a wet spring are consistent with seasonal soil movement rather than progressive failure. One reading cannot show you that. Two, six months apart, can.

05
Log floor elevations too

Cracks are the symptom; the shape of the floor is closer to the cause. Our elevation log records two readings over time, which is what makes it evidence rather than a number.

06
Hand the log to whoever assesses it

An engineer arriving to a house with eight months of dated readings can answer a question that an engineer arriving cold can only estimate. It also makes you a considerably harder person to oversell to.

The damage classification engineers actually publish

Burland’s classification of visible damage to walls is published by the International Society for Soil Mechanics and Geotechnical Engineering and was adopted with slight modifications by BRE, the Institution of Structural Engineers and the Institution of Civil Engineers.

It describes damage to masonry and plaster by what it takes to repair it, a more honest framing than "serious or not serious".

OptionWhat it means for you
0 — Negligible
less than about 0.1 mm
Hairline cracks.
1 — Very slight
up to 1 mm
Fine cracks easily treated during normal decoration. Damage generally restricted to internal wall finishes; close inspection may reveal some cracks in external masonry.
2 — Slight
up to 5 mm
Cracks easily filled; redecoration probably required. Cracks may be visible externally and some repointing may be needed for weathertightness. Doors and windows may stick slightly.
3 — Moderate
5 to 15 mm, or several greater than 3 mm
Cracks require opening up and patching by a mason. Repointing and possibly some brickwork replaced. Doors and windows sticking. Service pipes may fracture. Weathertightness often impaired.
4 — Severe
15 to 25 mm, and depends on the number of cracks
Extensive repair involving breaking out and replacing sections of wall, especially over doors and windows. Frames distorted, floor sloping noticeably, walls leaning or bulging noticeably, some loss of bearing in beams, service pipes disrupted.
5 — Very severe
greater than 25 mm, and depends on the number of cracks
Major repair involving partial or complete rebuilding. Beams lose bearing, walls lean badly and require shoring, windows broken with distortion, danger of instability.

Source: Mair, Taylor & Burland (1996), Prediction of ground movements and assessment of risk of building damage due to bored tunnelling, ISSMGE. Categories 0–2 are aesthetic, 3–4 affect serviceability, and 5 affects stability.

The paper calls "the division between damage categories 2 and 3" an important threshold: about 5 mm. The scale grades damage to walls and finishes, not foundation failure.

The caveat printed on the scale itself matters as much as the numbers.

The table carries this note in its own header: "Crack width is only one factor in assessing category of damage and should not be used on its own as a direct measure of it."

One 6 mm crack in a render coat is a different finding from twenty 3 mm cracks across a load-bearing wall with a door frame out of square.

What to ask before you sign

What do you think is causing this, and what evidence are you basing that on?
Which damage category would you put this in, and why?
Is there evidence this is active, or could it be historical?
Have you measured floor elevations, or only looked at the cracks?
What would you expect to see in six months if your explanation is right?
What would change your mind about the diagnosis?
Is there a cheaper thing I should do first, such as grading, downspouts or irrigation, before structural work?

The full checklist is on choosing a foundation contractor.

Where this goes wrong

Believing the direction rule

We could not find a single code, standard or engineering society document that assigns diagnostic meaning to crack orientation. See the FAQ below. It is the most repeated claim in the field and the least supported.

Treating one photograph as evidence

A photograph shows a crack. It cannot show movement, and movement is the question. Nobody — us, an engineer, or a contractor — can tell from an image whether something is progressing.

Filling the crack before recording it

Sealing it destroys the evidence and does nothing about the cause. If it is being filled for weathertightness, photograph and measure first.

Panicking about new-build shrinkage cracking

Concrete shrinks as it cures. Fine cracking in a slab or a foundation wall in the first couple of years is common and usually unrelated to settlement.

Getting the diagnosis from the seller of the cure

A free inspection can be entirely honest and is still a sales visit. It is also, in practice, where a category-2 crack most often becomes a piering proposal.

Assuming cosmetic means permanent

The classification describes damage today. A category-1 crack in an actively moving structure is a category-1 crack so far. That is exactly why the monitoring matters more than the measurement.

Get connected

Tell us about the work

Three short steps: the job, your home, and how to reach you. Nothing goes to a foundation company unless you agree.

A straight answer first. Matching is still being built, so this is not an instant quote. If a wall is moving or a door frame has racked out of square this week, call a structural engineer first.

Find a Foundation Professional

Tell us what you are seeing and where you are. We will follow up about connecting you with foundation professionals serving your area.

Step 1 of 3: your project

Questions

Is a horizontal crack in a basement wall always serious?
Not always. No code, standard or engineering society publication we found says horizontal means structural and vertical means harmless. The mechanics are real: a block wall pushed by soil tends to crack along a mortar joint. A horizontal crack with the wall bowed inward is serious, but the bowing is what makes it serious.
How wide does a foundation crack have to be to worry?
About 5 mm, or several cracks wider than 3 mm. The authors of the published damage scale call the step from category 2 to category 3 the important threshold. The same paper warns that width alone is not a direct measure of damage, so treat it as a guide, not a verdict.
Are stair-step cracks in brick a foundation problem?
Sometimes. A stair-step crack follows the mortar joints because they are the weak plane, so the pattern says more about masonry than cause. What matters is width, whether it is widening, and whether the wall has rotated or bulged. Brick veneer can also crack from its own supports and ties.
How much floor slope is too much?
A tilt over 1% is usually noticeable, according to the ASCE Texas Section guidelines, which note the ADA treats 2% as too large. Engineers judge foundation deflection against L/360, and discount deviations measured over less than about 20 feet, where built-in unevenness dominates.
Can I just monitor a foundation crack instead of paying someone?
Often, yes. Categories 0 to 2 on the published scale are aesthetic damage, and monitoring is what most cracks deserve. Do not wait when the structure itself has moved: sloping floors, leaning or bulging walls, frames out of square, or any sudden change.
Is there an American version of the crack damage scale?
Not that we could find. We searched ICC, ASTM, ASCE, ACI and federal sources and found no equivalent. US practice appears to use the British classification directly, or to work from deflection and tilt criteria instead of crack width.
What about the angular distortion limits quoted online?
Figures like 1/500, 1/300 and 1/150 are attributed to Skempton and MacDonald (1956) and Bjerrum (1963), both behind paywalls we did not read. The figure we can source is from Burland: local slope deviation over 1/100 is normally clearly visible, and overall deviation over 1/150 is undesirable.

Written and audited by

HyreFoundation Research Desk

Primary-source research and fact checking

We read the model building code, the federal soil and moisture guidance and the insurer position ourselves, and we publish each figure with the document it came from and the date we retrieved it.

Where a number cannot be traced to a primary source, we say so instead of printing it. HyreFoundation does not inspect, engineer or repair foundations. Authorship is organizational: this desk, not a named persona.

40
pages published on this site
12
free tools, each showing its assumptions
0
foundation repairs we perform or sell
0
paid placements or rankings

How this desk works

  • Primary sources only. Code statements come from the International Residential Code, Chapter 4, as the model code. Soil statements come from USDA and USGS. Insurance statements come from the policy language and the Insurance Information Institute. We open the source rather than cite a blog that cites it.
  • The code is a model, and adoption is local. Your jurisdiction adopts an edition and amends it, and foundation chapters are amended more than most because frost depth and soil are local facts. Every code figure says so.
  • We hold no job data. We have no dataset of bids, pier counts or invoices, so no page claims an average from one. Calculators show the arithmetic and the inputs they assumed, and label the result as an estimate.
  • We do not diagnose your house. A crack guide can tell you which pattern to measure and when to call a structural engineer. It cannot tell you whether your foundation is failing. An engineer on site does that work.
  • We do not repair foundations, and we take no payment for placement, ranking or a favorable mention. Nobody buys a position on this site.

Authorship on this site is organizational: the analysis belongs to the desk rather than to a named individual, and we do not publish credentials we do not hold. Our editorial policy sets out how we source, date and correct what we publish.

Sources & retrieval dates

Web Soil Survey — soil shrink-swell properties by address , USDA Natural Resources Conservation Service Retrieved Retrieved September 18, 2026.

Information on this page is general and varies with soil, climate, foundation type, building code, insurance policy and circumstance.

It is not engineering, insurance or legal advice, and no page can tell you whether your house has a structural problem: that needs a licensed professional on site.

HyreFoundation is an independent foundation resource, not a repair contractor or an engineering firm, and does not perform, supervise or warrant foundation work.