Grey belt has been read by a lot of landowners as a reclassification, as though a parcel is either grey or it is not, and the ones that are can be built on. The decisions do not support that reading.
What the appeals have actually turned on
Analysis of grey belt appeal decisions published by Urbanist Architecture in March 2026 found that where a decision principally turned on grey belt, 79% succeeded. That figure on its own has been quoted a great deal and it is misleading without the one underneath it.
The same analysis found that schemes passing the sustainable location test succeeded 72% of the time, and schemes failing it succeeded 5% of the time. Those two numbers are the whole story. Grey belt status gets a scheme into the argument. Sustainable location decides it.
The order these questions get asked in
Does the land genuinely fail to serve Green Belt purposes, is there evidenced unmet need, and is the location sustainable. Most refusals we have read fall at the third, having comfortably cleared the first two. A site that is grey belt and badly connected is still a refusal.
What the August framework changed
Two things. A five-year supply shortfall, or a Housing Delivery Test result below 75%, now counts as evidenced unmet housing need under S5(1)(j). And the definition of grey belt dropped the footnote 7 exclusion, so land that was previously excluded because another protected policy would restrict development can now be classified as grey belt.
That second change is easy to misread as generosity. Land can now carry the label while heritage, habitat or flood constraints still stop anything being built on it. The classification became broader. What you can do with the classification did not.
Neither change touches the order the questions get asked in. They make the second one easier to answer in more places, which mostly means more schemes reaching the sustainable location test and failing there instead.
The golden rules are doing quiet damage
The Golden Rules sit at GB8 and require affordable housing 15 percentage points above the highest existing local policy requirement, capped at 50%. That is a viability problem before it is a policy problem, and it shows up as a refusal ground in its own right where applicants have argued their way down from it.
A grey belt case that arrives with a viability argument attached is a weaker case than one that does not. That is worth knowing before the scheme is designed, rather than after.
Sources
- Urbanist Architecture, Examples of grey belt planning and appeal decisions, March 2026 update. Success rates and sustainable location figures are theirs.
- Ministry of Housing, Communities and Local Government, National Planning Policy Framework, published 17 August 2026, on S5(1)(j) and the GB8 Golden Rules.
- Planning Geek, NPPF 2024 vs 2026, on the removal of the footnote 7 exclusion from the grey belt definition.