Quantification of assemblages of ceramic building material (CBM) is now normal excavation practice, but there is insufficient standardisation of methodology and little study of the validity or uses of these analyses. Consequently, it is difficult to interpret confidently the resulting analyses or make comparisons between sites. This paper suggests a simple approach and tests its validity empirically in urban environments where the high level of constructional churn should generate assemblage homogeneity. It then shows how the results can help with site and building interpretation.