Some studies have suggested that the prevalence of atopic eczema may vary between geographical regions. This descriptive study investigates the regional and subregional geography of reported and examined eczema prevalence at the age of 7, 11 and 16 years in Britain using data from the 1958 birth cohort study (n = 828). Estimates of the relative risk of reported eczema associated with residence in each region of the country were calculated and the regional distribution of reported and examined eczema prevalence was compared. The reported prevalence of eczema was mapped at the smaller county level. Comparisons were made with the county-level distribution of asthma and hay fever prevalence. The study showed a marked and statistically significant variation in eczema prevalence across the regions in Britain which was present for examined as well as reported eczema. The highest risk was associated with four regions: North Midlands; Eastern; London and the South- East; and Southern. This regional pattern was not altered significantly after adjustment for social class and family size. The geographical distribution of eczema prevalence was largely maintained when analysed at the county level. Few similarities were found between the county-level distribution of eczema prevalence and that for asthma and hay fever. Explanations for this strong regional variation now need to be sought in terms of environmental and life- style associations.