Privacy distances between windows and the 21 metre convention

The 21 metre privacy distance planning convention is a design guide figure, not a rule of law. It says that habitable room windows facing each other across private space should be about 21 metres apart, with roughly half that from a window to a boundary or a blank wall. No statute contains it, the National Planning Policy Framework does not mention it, and a scheme can comply with it and still be refused on overlooking.

Knowing where the number sits in the hierarchy changes how you design to it. The question an officer is answering is whether future and existing occupiers have a reasonable standard of privacy. The 21 metres is a shorthand for the usual answer in a suburban two storey context, and it stops being the answer the moment the context changes.

Where the convention came from

The separation distance is usually traced to interwar housing layout guidance, which recommended something close to 70 feet between facing rows of two storey terraces. The reasoning at the time was as much about daylight and sunlight into low-rise terraces as about privacy, and the figure converted neatly into 21 metres when units changed.

The provenance is contested, and nobody should present it as a precise lineage. What is clear is that the number entered local design guidance and stayed there, long after the building types it was derived from stopped being what gets built.

That history explains the convention's main weakness. It was calibrated for two storey houses with back gardens facing each other across a block. Applied to a three storey townhouse, a flatted block with balconies, or a mews arrangement, it either over-controls or under-controls, and officers know this.

How design guides express the standard

Councils write the standard into householder design guides, residential design supplementary planning documents and sometimes into local plan policy. The wording varies, and the variation matters.

Read the local document rather than relying on the headline number. Some guides apply the 21 metres only where both windows serve habitable rooms. Some treat kitchens as habitable and some do not. Some measure to the boundary and some to the nearest part of the building. A scheme designed to the wrong variant of the standard fails a test it was trying to pass.

Oblique angles and partial overlooking

Direct facing windows are the worst case, and almost no real site presents it. Once the relationship is angled, the view from one window into the other becomes partial, and most guides allow a reduced distance.

The practical measure is the angle between the two window planes and the sight line between them. Where the angle is wide enough that an occupant has to stand at the glass and look sideways to see into the other room, officers generally accept a shorter separation. Where the sight line runs straight into the centre of a facing window, the full distance applies even if the buildings are not parallel.

Two drawings settle most of these arguments. A plan with sight lines drawn from each affected window at the angle a person could actually see, and a section showing the floor levels. Without them, the discussion becomes two people asserting different interpretations of the same geometry.

Ground floor against upper floor windows

Ground floor to ground floor overlooking is usually controlled by boundary treatment rather than by distance. A 1.8 m fence or wall removes it, which is why the convention is really about first floor and above.

Level differences change the picture in both directions. A proposal on ground that sits 1.5 m higher than the neighbour gains an effective storey of elevation, and a first floor window behaves like a second floor one. Sloping sites therefore need sections rather than plans to assess, and the common error is to measure the horizontal distance and treat the matter as closed.

Dormers are the other recurring issue in householder work. A rear dormer places a habitable room window higher and often closer to the rear boundary than the original elevation, and the angle of view down into a neighbouring garden is what draws objections. Setting the dormer back from the eaves and reducing its width does more for the privacy case than reducing its height.

Obscure glazing, fixed lights and screening

Where distance cannot be achieved, the standard toolkit is well established and officers are familiar with every part of it.

• Obscure glazing to a stated obscurity level, usually to at least 1.7 m above internal finished floor level.
• Windows fixed shut below that height, with opening lights above it.
• High level windows serving the room with a cill above eye level.
• Angled or fin screens on balconies, with the screen specified rather than described.
• Window reveals deep enough to limit the view angle, which works on contemporary elevations and looks contrived on traditional ones.
• Planting, which officers discount almost entirely because it is seasonal and can be removed.

Two cautions apply. Obscure glazing to a habitable room's only window is usually resisted, because it solves the neighbour's privacy by removing the new occupier's outlook, and the approach to daylight in single aspect flats shows why that trade rarely holds up. And a condition requiring obscure glazing to be retained in perpetuity is enforceable, so treat it as a permanent design decision rather than a drafting fix.

Dense urban sites where the distance is relaxed

In inner urban areas the 21 metres is routinely unachievable and routinely relaxed. Local plans in high density boroughs often say so expressly, substituting a qualitative test about overlooking, outlook and the design response.

What replaces the number is an argument about the prevailing pattern. Where the existing streets already have facing windows at 14 metres, a new building at 15 metres is not introducing a new condition. Where a scheme would create a relationship tighter than anything nearby, the burden shifts to the applicant to show the design handles it.

This is also where the amenity assessment becomes a package rather than a single figure. Privacy, daylight, sunlight and outlook are assessed together, and a tight separation with good daylight and a considered window arrangement reads better than a compliant separation with poor light. How local plan amenity policies use BRE 209 covers the daylight side of that package.

Evidencing acceptable privacy without the distance

When the distance is not met, the application has to do the work the number would otherwise have done. Four pieces of evidence carry most of the weight.

A measured plan showing the separation at its narrowest point, with sight lines from every affected window, and the distance stated rather than left to be scaled. Sections through the critical relationship showing floor levels and cill heights on both sides. A window by window schedule saying what each window serves and how it is treated. And a short written assessment against the specific wording of the local design guide, naming the policy and the paragraph.

Where daylight is also in question, a Daylight and Sunlight Assessment covering the same windows lets the officer consider both in one place. Where the issue is sunlight on gardens rather than windows, an Overshadowing Study is the right document.

The honest version of this submission concedes where it falls short. An officer who finds an unacknowledged 16 metre relationship in your own drawings stops trusting the rest of the document, and that costs more than the shortfall would have.

You can see which reports a scheme with amenity issues is likely to need and draft the amenity sections from your own site details before commissioning the modelling.

This is general guidance, not legal advice. Separation standards are set locally and worded differently in each authority, so check your council's design guidance or take advice from a qualified professional before relying on it. PlanScope provides AI-assisted draft reports for planning applications. Reports should be reviewed before submission. We do not guarantee planning approval.