Who this guide is for
- Owners considering removing or opening up an internal wall
- Buyers trying to understand what an open-plan alteration would involve
- Anyone who has been given conflicting opinions about the same wall
- Owners of an older building with no record of previous alterations
- Anyone preparing for a first conversation with a structural engineer
Three tests that settle nothing
Thickness is the first. The masonry entry is direct that a thick wall is not automatically load-bearing and a thin one is not automatically a partition, and only the load path settles it.
Material is the second. A framed wall may be carrying floors and roofs, and appearance does not distinguish it from one that is not; in a light steel building the sections themselves look the same, which is why which walls are carrying has to be readable from the drawings rather than by inspection.
Reaching the ceiling is the third, and it fails in both directions. A partition that meets the ceiling may only be dividing space, and a wall that stops at a suspended ceiling may still be doing something above it.
There are three questions, not one
The reference entries separate duties that everyday conversation runs together. A wall may be carrying vertical load, it may be providing restraint to another element, and it may be part of the arrangement resisting horizontal action. Those are different jobs and a wall may do one, two or all three.
The stability entry is explicit that a wall is frequently judged loadbearing or not by what sits above it, ignoring the quite separate question of horizontal action, and that the distinction between a partition and a stability wall is not thickness or material but whether the element is connected into the path.
- Is it carrying floor or roof load down to something beneath?
- Is it restraining a wall, a floor or a frame that relies on it?
- Is it part of the arrangement resisting wind and other horizontal action?
- Is anything above it bearing on it that was added later?
The duty that leaves no trace above it
This is the case that catches people out. A shear wall entry warns that a wall carrying no floor may still be the building lateral stability, and that its removal would not be visible as a sagging floor. A masonry wall at an angle to another keeps that wall in line, and the floors that bear on a wall also hold it in position.
So the absence of anything obviously sitting on a wall is not evidence. Restraint runs in the opposite direction to load: cut a floor away to form a void, or take out a returning wall to open a plan, and the wall loses what was keeping it in line even though nothing about its vertical load has changed.
What can be observed, and what observation cannot do
Some things are worth noting because they help whoever assesses the building. Which way the floor above spans, where joists and beams land, where chimney breasts and returning walls sit, and what the wall is built from are all useful observations.
They are not a conclusion. The masonry entry says assumptions made from appearance are unreliable, and the stability entry says elements resisting horizontal action are frequently indistinguishable from ordinary walls and floors once finishes are in place. Observation narrows the question; it does not answer it.
A partition can behave like a support anyway
Even a wall that was never designed to carry can end up carrying. A partition built tight to a soffit becomes a prop: floors and beams deflect under load and go on deflecting over time, and a partition standing in the way takes load it was never arranged for, with studs bowing, boards cracking along the head line and doors binding.
The stability entry makes the same point more generally: non-structural things fixed to the path join it whether or not that was intended, so a partition built tight against a wall attracts force it was never designed for. Cracking along the top of a partition is commonly the visible result of that rather than a decorating problem.
What is inside the wall is part of the answer
A framed partition is mostly invisible once boarded. The cavity may be empty, filled, crossed by a resilient layer, carrying services, or lined with concealed backing, and none of those choices changes the room appearance while all of them change what the wall can do and what can be done to it.
That is why the entries treat the record made at boarding stage as behaving like a component. What a later fixing reaches behind the face differs between a backed area and an unbacked one, and nothing on the finished face distinguishes them.
The building history is evidence
Older buildings rarely carry a record of what was altered and when. Establishing which walls are carrying, what has already been removed and what is currently propping what is a survey exercise, and it matters because a previous alteration may sit on the load path being relied on now.
The masonry entry puts that directly to professionals as a question: have previous alterations been identified, and do any of them affect what is being proposed here.
How it actually gets established
In a new or recent building it is a drawings question, because which walls are carrying and bracing has to be readable from the information rather than from the wall. In an older building it is an investigation: a survey, opening up where necessary, and an engineer describing how they established the answer rather than simply giving it.
The useful question to ask, taken straight from the reference entries, is not whether a wall is loadbearing. It is: can you confirm which of these walls are load-bearing, and describe how you established that.
Before asking about a wall
- 1Note which direction the floor above appears to span, and where anything lands on the wall.
- 2Note whether the wall has returning walls at an angle to it.
- 3Note whether the wall runs through more than one storey and what sits above it.
- 4Note whether the wall stops at a ceiling line or continues to the structure above.
- 5Note whether the wall is built from the same materials as ones you know to be structural.
- 6Photograph both faces, the junctions at each end and the head condition.
- 7Gather any drawings, previous approvals or alteration records that exist.
- 8Ask the previous owner or neighbours what has already been changed.
- 9Ask specifically about horizontal action, not only about what sits above.
- 10Ask what restrains the wall at present and what would restrain it afterwards.
- 11Ask how the answer was established, rather than only what it is.
- 12Record the answer and keep it with the building documents.
Common mistakes to avoid
- Reading thickness as the answer, when only the load path settles it.
- Treating a wall that reaches the ceiling as carrying, when in a framed building it may only be dividing space.
- Assuming a wall with nothing obviously above it can go, when it may be part of the horizontal path.
- Forgetting that floors and returning walls restrain a wall as well as load it.
- Cutting a floor opening or removing a return and treating that as unrelated to the walls around it.
- Judging structural studwork by appearance when the sections look the same as non-structural ones.
- Reading cracking along the top of a partition as decoration rather than as the head being loaded.
- Relying on a verbal opinion without asking how the answer was established.
When to involve a professional
- Nothing here establishes whether a particular wall may be opened, altered or removed; that is a determination for a structural engineer on the specific building.
- No adequacy for any horizontal action is stated or implied, and that assessment belongs to a qualified engineer for the specific building.
- Elements resisting horizontal action are frequently indistinguishable from ordinary walls and floors once finishes are in place.
- Evidence of movement in a building should be assessed before any alteration is designed rather than diagnosed from a description.
- Shared and party structures carry legal obligations in many jurisdictions, and those sit outside the scope of this reference.
Frequently asked questions
Questions readers ask about this topic
Does a thicker wall mean it is loadbearing?
No. The reference entries are explicit that a thick wall is not automatically load-bearing and a thin one is not automatically a partition, because only the load path settles it. Thickness can follow from acoustic, fire, service or historical reasons instead.
If nothing sits on a wall, can it be removed?
Not on that basis alone. A wall carrying no floor may still be part of the arrangement resisting horizontal action, or may be restraining another wall or a floor, and its removal would not be visible as a sagging floor.
Why does reaching the ceiling not settle it?
Because it fails both ways. A partition that meets the ceiling may only be dividing space, and a wall stopped at a suspended ceiling line can still be doing work above it. The head condition describes how the wall was built rather than what it carries.
My partition is cracking at the top. Is it structural?
Cracking along the head is commonly the visible result of a partition restraining, or being loaded by, the structure above it. A partition built tight to a soffit that deflects becomes a prop, which is a movement problem at the head rather than proof the wall is carrying.
What should I ask an engineer?
The entries suggest asking which walls are load-bearing and how that was established, what restrains each wall now and afterwards, whether any wall is part of the horizontal arrangement, and whether previous alterations have been identified.
Keep reading