The causative agents of variant Creutzfeldt-Jakob disease (vCJD) and of bovine spongiform encephalopathy (BSE) are currently indistinguishable. However, the route(s) by which humans became infected remain unknown. The path by which humans were infected with the BSE agent might impact on the geographical distribution of cases and we therefore sought evidence of regional variation and local clustering of vCJD cases. With the notable exception of a group of five cases in Leicestershire, the absence of local clustering of vCJD cases is compatible with most human exposure to the vCJD agent arising through routes that result in the risk of infection being similar over wide geographical areas, rather than through small-scale local events. Infection through the consumption of mechanically recovered meat (MRM) contaminated with the BSE agent was a potential route for such widespread exposure. An ecological analysis relating the regional incidence of vCJD to historical dietary data does not provide a dear evidence that humans became infected through consumption of MRM.