Have we learned to count yet? A health services perspective on the 2001 census

English health services require reliable population denominators to guide equitable sharing of resources. 2001 census counts for England are compared with the corresponding 2000-base projections for 2001, both at the national scale and for individual local authorities and also with corresponding 1996-base projections. Comparison is made with the 2001 census for Northern Ireland. National patterns of variation predominantly may be explained by mis-estimation of international out-migration, but local patterns of variation may be more specifically related to mis-estimation of international in-migration. This suggests two distinct patterns of systematic error in official UK migration statistics between 1991 and 2001.