Subsidence is downward movement of the ground beneath a building's foundations. A structural engineer diagnoses the cause, distinguishes it from settlement or thermal cracking, and specifies the repair. Underpinning is only one possible outcome, and it is often not the right one.
Most cracks in a London house are not subsidence. But telling the difference matters, because the wrong diagnosis costs money in both directions: an unnecessary underpinning scheme, or a real ground problem left to get worse for another two summers.
Diagnosis is a structural engineering exercise, and it starts with the crack pattern rather than the crack width.
What is subsidence, and what is it not?
Subsidence is downward movement of the ground supporting your foundations, caused by the ground itself changing rather than by the weight of the building.
Settlement is downward movement caused by the building compressing the ground beneath it. Almost every building settles. It usually happens early in the building's life, and it usually stops.
Heave is upward movement, most often when clay soil re-wets and swells after a tree is removed or a leak is repaired.
Thermal and moisture movement is the everyday expansion and contraction of materials, producing fine, often seasonal cracking that opens in winter and closes in summer.
The word subsidence carries real financial consequences. It is a defined insured peril, it has to be disclosed on sale, and it affects future insurability. Applying it to a crack that is actually thermal movement is not a harmless error.
How common is subsidence in the UK?
Common enough to be a significant claims category, and the cost per claim is rising.
According to the Association of British Insurers, subsidence-related claims totalled £153 million in the first half of 2025, with insurers supporting almost 9,000 households and an average payout of £17,264 per claim. The ABI linked the spike to an exceptionally warm and sunny spring, the warmest on Met Office record (ABI).
The trend has continued. By the first quarter of 2026 the average household subsidence claim had risen 9 per cent year on year, from £16,295 to £17,820 (ABI).
The pattern behind those figures is clay. Hot, dry summers shrink clay soils, clay-rich areas of London and the south east are disproportionately affected, and the ABI has been explicit that climate change is increasing the risk.
What do crack patterns actually tell you?
A structural engineer reads the pattern, not just the width. The ABI's own guidance points homeowners at cracks more than 3 millimetres wide, roughly the depth of a £1 coin, that run diagonally, are wider at the top than the bottom, and are visible both inside and outside the property. Here is what each of those features is telling you.
Direction. Subsidence cracks are typically diagonal, running from the corners of openings, because the wall is rotating about a point of restraint. Vertical cracks of consistent width along a whole storey are more often thermal, or a wall tie issue.
Taper. Wider at one end than the other reflects rotation rather than a uniform pull.
Continuity. Movement cracks pass through more than one material. They carry through brickwork, render and internal plaster, and usually show on both faces of a wall. A crack confined to plaster is decoration, not structure.
Location. Corners of window and door openings, the junction between a bay and the main wall, and the connection between an extension and the original house. Those are stress concentrations and stiffness changes, so they crack first.
Speed of onset. The ABI notes that subsidence cracks usually appear very suddenly rather than gradually. Other indicators include doors and windows sticking without an obvious cause, or wallpaper ripping in areas with no damp.
Season. Clay shrinkage cracks open in late summer and autumn and close over winter. That seasonal signature is diagnostic in itself, which is why monitoring across a full cycle is so useful.
What causes subsidence in London homes?
Clay shrinkage. The dominant cause in London and the south east. Clay soils lose water in dry weather and shrink, and foundations sitting on them move down. Shallow Victorian and Edwardian foundations, common across inner London, are particularly exposed.
Trees and vegetation. Roots extract water from clay, amplifying shrinkage. Distance, species and water demand all matter. Removing a mature tree next to a house can cause heave, so removal needs engineering input rather than enthusiasm.
Leaking drains. A leak washes fines out of the ground and softens it locally. This is one of the more common causes, one of the cheapest to fix, and one of the most frequently missed. A CCTV drainage survey is standard early evidence.
Made ground and historic fill. Large parts of London sit on made ground, including bomb damage backfill, old watercourses, quarries and infilled basements. Fill compresses unpredictably.
Nearby excavation. Basement construction next door, deep foundations or tunnelling can all change ground conditions.
Inadequate original foundations. Some Victorian foundations are shallow, and some later extensions were founded at a different depth from the house they were attached to, which produces differential movement at the junction.
What does a structural engineer do on a subsidence investigation?
The process is deliberately evidence-led, because the repair is expensive and the diagnosis drives it.
- Desk study. Geological maps, historic mapping, previous claims history, tree records.
- Visual inspection. Crack survey, level survey of floors and thresholds, distortion survey of the elevation, examination of drainage and vegetation.
- Monitoring. Crack gauges or precise level monitoring across at least one full seasonal cycle, where the movement is not clearly progressive.
- Drainage survey. CCTV inspection to eliminate or confirm leakage.
- Site investigation where justified. Trial pits to establish actual foundation depth and type, soil sampling, root identification.
- Diagnosis and report. A written structural report stating the cause, whether movement is active, and the recommended remedial approach.
- Design of the remedy, with drawings and calculations where structural works are needed.
Steps three to five are the ones homeowners want to skip and should not. Underpinning designed without knowing the existing foundation depth, or the cause of movement, is a guess with a five figure price tag attached.
What should the engineer's report actually contain?
A subsidence report is a document you may end up relying on with an insurer, a neighbour, a lender or a buyer, so vagueness in it is expensive. A useful report states:
- What was inspected and what was not. Including anything that could not be accessed, because that limits the conclusions
- The crack pattern as measured, with widths at recorded positions and dated photographs, not general description
- The cause, stated plainly, and whether the movement is active, slowing or historic and stable
- The evidence behind that conclusion, so a third party can follow the reasoning rather than take it on trust
- What was ruled out and why, which is often as valuable as the finding itself
- The recommended remedy, with the cheaper options addressed explicitly rather than passed over
- What happens next, including any monitoring period and what would change the recommendation
If a report recommends underpinning without stating the cause of movement, the existing foundation depth, and why removing the cause would not be sufficient, it has not done the work. Ask for those three things before anybody prices a scheme.
When is underpinning the right answer?
Underpinning extends or replaces the existing foundation so it bears on deeper, more stable ground. It is effective, disruptive and expensive, and it is the correct answer less often than homeowners expect.
Underpinning is usually appropriate when:
- movement is active and progressive rather than historic
- the cause is a genuine loss of bearing capacity that cannot be removed
- the foundation is demonstrably inadequate for the ground beneath it
- cheaper interventions have been eliminated by evidence, not by assumption
Underpinning is often not the answer when:
- the cause is a leaking drain, which can simply be repaired
- the cause is a tree whose water demand can be managed
- movement is historic and stable, in which case repair the cracks and monitor
- only part of the building is affected, and partial underpinning would create a stiffness change and a new differential movement problem
That last point is where cheap schemes cause expensive damage. Underpinning one section of a building creates a hard point next to a soft one. Where partial underpinning is genuinely required, the transition has to be designed. Removing the cause and monitoring, sometimes with crack repair and localized strengthening, resolves a substantial share of cases, and that is a better outcome than underpinning.
What about insurance, party walls and building control?
Insurance. Subsidence is a standard peril on buildings insurance, usually with a higher excess. If you suspect subsidence, notify your insurer before commissioning major works. Insurers normally appoint their own engineer. You are entitled to your own independent structural engineer's opinion, and where the diagnosis is contested that opinion is worth having.
Party walls. Underpinning a party wall, or excavating within three metres of a neighbour's structure to a depth below their foundations, engages the Party Wall etc. Act 1996. Notice must be served and the timescales are real.
Building control. Underpinning is notifiable building work and requires Building Control approval.
A sensible sequence for a homeowner:
- Photograph and date the cracks, recording widths at fixed points
- Check the obvious: blocked or leaking gutters, downpipes discharging at the wall, overgrown vegetation
- Get a drainage survey
- Notify your insurer if the pattern suggests structural movement
- Instruct a structural engineer for an independent diagnosis
- Do not commission underpinning from a contractor who has not seen a diagnosis
That last point saves the most money.
Subsidence, settlement, heave or thermal movement?
| Factor | Subsidence | Settlement | Heave | Thermal |
|---|---|---|---|---|
| Direction | Downward | Downward | Upward | Cyclical |
| Cause | Ground changes beneath foundation | Building compresses ground | Ground swells, often clay re-wetting | Material expansion and contraction |
| Timing | Often after dry summers | Early in life, then stops | After tree removal or leak repair | Seasonal, repeating |
| Insured peril | Usually yes | Usually not | Usually yes | No |
If you are planning structural alterations rather than diagnosing movement, see our guides to load bearing wall removal and extensions and loft conversions.
Frequently Asked Questions
Written by the founder of PV Consulting, a Chartered Engineer (CEng MICE, Institution of Civil Engineers) with over 30 years of experience in structural and civil engineering, including residential structural assessment across London and the south east. PV Consulting Ltd, company number 08294917, registered in England and Wales, professional indemnity insured.


