Trees and Subsidence: Prune, Fell or Underpin?

How a tree moves a house

Clay is the only common soil that shrinks as it dries and swells as it wets. A tree's roots take far more water out of the ground in summer than the rain puts back, so the clay beneath and around it shrinks through the growing season and recovers through the winter. A house whose foundations sit within that zone rides up and down with it, and because the drying is greatest near the tree and least far from it, the movement is uneven and the walls crack. The stresses a drying clay develops are many times the weight of a house on its footings, which is why a tree can move a building that has stood still for a century.

What governs the tree's effect is its leaf area, the size and density of the crown, rather than its height, and the depth to which its particular species will dry the ground. Oak, poplar, willow, elm and the cypress family are the species that dry clay deeply and can build up, over decades, a zone that never fully rewets in winter. That persistent dry zone is the crux of every decision that follows. Our guide to foundations near trees on clay sets out which species matter and how far their influence has been recorded; the short version is that oak, poplar and willow have been implicated up to 18 to 20 m away, most other large trees up to about 7 to 12 m, and small garden trees rarely beyond 3 to 9 m, with the caveat that these are records of where damage was found, not distances at which damage is likely.

Proving it is the tree

A tree within the recorded distance is a reason to investigate, not a diagnosis. Cracks near a tree can equally be a leaking drain by the kitchen, a failed lintel or thermal movement, and blaming the tree without evidence leads to felling that solves nothing or, worse, starts heave. The evidence an engineer looks for:

• A seasonal pattern. Level and crack monitoring that goes down through the summer and part of the way back up through the winter is the signature of clay and vegetation. A leaking drain goes one way regardless of the month.
• Where the movement is. The distortion is greatest nearest the tree; a small shrub close to the wall produces local movement, a large tree further away produces widespread movement.
• Roots beneath the footing. A trial pit beside the affected foundation records the depth of the footing and whether live roots pass below it. Roots can be identified to species in a laboratory, and by DNA where several trees are candidates, though a root's presence shows the tree reaches the spot, not on its own that it is drying it.
• The moisture profile. Samples from the trial pit or a hand auger, taken every 200 to 250 mm down, show how dry the clay is against what it would hold at field capacity and against a borehole away from the tree. Soil suction tests are the most reliable measure of desiccation. Autumn boreholes show the total drying; spring boreholes show what has failed to rewet, the persistent deficit.
• The tree itself. An arboriculturist assesses the species, crown volume, vigour, previous pruning and history. The tree's crown volume is a better guide to its water demand than its height, and a tree that has been recently reduced or pollarded may be drawing far less than its size suggests.

Seasonal or persistent: the finding that decides everything

If the clay under the tree rewets fully each winter, the tree's effect is seasonal, and reducing its water demand reduces the amplitude of the movement; remove the tree altogether and the ground simply recovers over the following year or two. If the tree has dried the clay deeply for years and the winter rain no longer reaches the bottom of the dry zone, the deficit is persistent, and any large reduction in the tree's demand, let alone felling, starts a rehydration that can lift the house for many years. There is no way to hold the clay at a moisture content where it neither shrinks nor swells. That is why the investigation is worth its cost, and why the question "shall we just cut it down?" is the wrong one to start with. Our guide to heave vs subsidence covers what happens when it goes wrong.

Option one: manage the tree

Where the deficit is seasonal, the remedy is to reduce the tree's water demand enough that the movement falls within what the building can tolerate, and to keep it there. The method that works is crown reduction: taking the overall volume of the crown down by a specified proportion and reshaping it so that less air moves through the foliage, which is what research since 2004 has shown actually reduces transpiration. Pollarding, the hard shortening of the main branches, is the version used on street trees under a regular regime. Crown thinning, removing inner branches while keeping the crown's size, does not reduce water uptake and is no longer recommended for this purpose, whatever a tree surgeon may offer.

The catch is regrowth. Most trees respond to pruning by growing back vigorously, restoring their leaf area within a few years, so the reduction has to be repeated every two to three years for as long as the tree stands, and after any insurance claim closes that is the owner's cost and responsibility. The severity of each reduction is a judgement for an arboriculturist: enough that by the time the tree needs pruning again its crown is still well below the size that caused the damage, not so much that it ruins the tree or provokes a burst of juvenile growth. Species matters; London plane takes heavy pruning well and is rarely implicated in damage, while some others tolerate only light work. Trees nearby that are not implicated need no pruning at all. Level monitoring continues through at least one summer after the first reduction to prove it has worked, and the markers are left in for spring and autumn readings while the regime continues.

Root barriers are occasionally proposed as an alternative. They are rarely worth it: roots have been recorded passing beneath barriers nearly 3 m deep and over the top of shallower ones, the trench to install one cuts the roots that hold the tree up, and the barrier can change the way water moves in the ground. Root pruning and soil moisture control systems have similar problems.

Option two: fell the tree

Felling stops all water uptake at once, which sounds decisive, and for a low-value tree with a seasonal deficit it is often the sensible answer: the ground recovers by the following spring if the tree comes down in autumn, or immediately if it comes down in spring, and the cracks frequently close on their own. The problems arise where the tree has a persistent deficit beneath it. Then felling begins a rehydration that can lift the house for years, and where the house was built after the tree had already dried the ground, lifts it above the level it was built at. Felling a tree that predates the house is therefore never done on a hunch. The other costs of felling are the tree's own value, shade, cooling, wildlife and the street, which councils, insurers and a growing body of case law now weigh, and, for anything larger than a garden tree, the possibility of a felling licence being needed.

Option three: keep the tree and underpin

Where the tree is valuable, protected or simply not yours, and the deficit is persistent, the engineering answer is to carry the house past the problem. That means underpinning designed for a clay that may later swell: mini-piles taken below the zone the tree influences, designed to resist the uplift a rehydrating clay would exert, sleeved over their upper length, with ground beams on a compressible void former and a suspended ground floor where needed. It is the most expensive of the three options and the most disruptive, but it is the only one that is indifferent to what the tree does next, and it is sometimes what an insurer prefers to a felling that would generate a heave claim. For a garden wall, a garage or an outbuilding damaged by a valuable tree, there is a cheaper engineering alternative: bridge the wall over the roots on a lintel or a fence panel and leave the tree alone.

Before anyone touches the tree: the law

Check with the council whether the tree is covered by a Tree Preservation Order or stands in a conservation area. If it does, any work, to the crown or the roots, needs the council's consent first, and unauthorised work is a criminal offence. The application is made with the engineering and arboricultural evidence that the tree is causing damage; where the tree is a legal nuisance an exemption may apply, but that is a legal judgement, not a homeowner's. If the council refuses consent for work to a tree proven to be causing damage, the law provides for compensation for the loss that follows, typically the additional cost of the alternative repair. A felling licence from the Forestry Commission is needed where more than a small volume of timber is felled in a quarter, which can catch a row of large trees.

Where the tree belongs to a neighbour or the council, you cannot simply require its removal. The remedy is a claim in nuisance or negligence, and it turns not only on whether the tree caused the damage but on whether the owner could reasonably have foreseen the risk; the damages must be reasonable, and courts have been reluctant to order a tree felled where insurance would fund an alternative repair. Put the owner and their insurer on notice as early as possible, with evidence, because delay weakens the claim and the insurers' agreement on these cases applies only to domestic owners. For council trees there is an industry protocol that sets out the timetable and the evidence both sides exchange in the mitigation phase. Where access to a neighbour's land is needed for works to a tree, a hedge or a drain, the Access to Neighbouring Land Act 1992 provides for a court order if consent is refused. Your own insurer, having paid for the repairs, may pursue the tree owner in your name to recover its costs.

Prevention, and what not to do

There is no economic case for pruning every tree near every house in case it causes subsidence one day; the cost would dwarf the repairs it prevented. There is a good case for targeted, cyclical crown reduction in streets and gardens where damage has already occurred on clay, and for choosing low water demand species when planting near a house. Do not plant a poplar, willow or oak within its recorded influence distance of a house on clay, do not let a Leylandii hedge grow into a row of trees, and do not remove a mature tree near a house, or let a contractor do it, without an engineer's and an arboriculturist's view. If you are planning an extension near a tree, its foundations are designed for the tree and the clay from the outset, which our guide to foundations near trees explains.

Who pays

Under an accepted subsidence claim, insurers generally pay for the reasonable tree works that form part of stabilising the building, and for the monitoring that proves they worked, even though tree management is strictly not an insured cost; the repeat pruning after the claim is yours. They very rarely pay for work to a tree they do not insure. Our guides to subsidence insurance claims and subsidence repair costs cover the rest.

Where EMA Structures fits

We investigate tree-related movement by evidence, in stages, each priced before it starts. The subsidence assessment, £495 to £895 + VAT depending on the size of the property and the type of movement, covers the inspection, a record of every crack, a level survey along the brick courses, a diagnosis of the likely cause and a written report. Where the readings need time, crack and level monitoring is £1,200 to £2,000 + VAT for the set-up and three seasonal visits, £350 + VAT for each further visit. Trial pits, soil sampling, root identification and an arboriculturist's report are not included in those fees; they are quoted separately once the assessment shows which of them the case needs, and our specialist site team carries out the pits. Then we tell you which of the three options the evidence supports, design and specify the repairs or the anti-heave underpinning if that is the answer, handle Building Control and the paperwork, and certify the work. Start with a subsidence assessment, or send us photographs of the cracks and the tree.

General guidance on UK practice as at September 2026. Tree distances quoted are historical records, not safe distances, and legal remedies depend on the facts of each case; take legal advice before pursuing a tree owner or a council. Fees exclude VAT.