This paper investigates the relationship between premature mortality and material deprivation, and the differences in this relationship between urban and rural areas. We examine, given comparable measures of affluence or deprivation, whether residual differences exist between urban and rural areas for all-causes of death and, separately, for cancers, circulatory and respiratory diseases. Using 1990-92 mortality data for the 908 wards of Wales we apply statistical analyses based on tabular data and parametric Poisson regression models. Contrasts are sought between six urban and rural categories defined in terms of settlement sizes and the employment structure of rural areas. Inequalities in all-cause premature mortality are widest in the cities, narrowest in the deeper rural areas, and of intermediate and comparable value in other areas of Wales. This is largely a reflection of the different distributions of material deprivation in these areas. After controlling for differences in socio-economic characteristics, using deprivation measures, the tendency for lower mortality in deeper rural areas is substantially reduced. Residual mortality differences between urban and rural areas are shown to be dependent on the way deprivation is measured and the disease group under study. For cancers there are no residual mortality differences, while for respiratory and circulatory diseases some of the residual variation can be accounted for by employment variables, particularly previous employment in the coal mining industry. Copyright (C) 2000 Elsevier Science Ltd.