British historical sources on occupational information such as censuses, parish registers, and probate records describe many men with the unhelpfully vague term of ‘labourer’. This paper introduces a new method to allocate these labourers to occupational (sub-)sectors, a prerequisite for creating comprehensive and accurate historical occupational structures. The new method leads to a significant correction on the allocation shares used in the national accounts literature. Its results at national level are largely in agreement with another new approach, developed by Osamu Saito and Leigh ShawTaylor. But it has an important advantage over that approach: it is capable of allocating labourers at all geographic levels, and can thus generate local and regional occupational estimates, rather than only national ones.