OBJECTIVE To analyse the relation of ethnicity, social deprivation (Underprivileged area, UPA-score), social class V, unemployment and overcrowding on age- and sex-standardized mortality ratio (SMR). METHODS This ecological study used the SMR for people under the age of 65 years as dependent variable. Mortality data for 1983, classified by total population of country of birth (New Commonwealth and Pakistan), unskilled (social class V), unemployed, and overcrowded (>1/room) population with data from 1981 censuses. Underprivileged Area (UPA) score was applied in this study, an index widely used in the UK to identify underprivileged residential areas, calculated from information combining eight variables derived from the decennial census. The data were analysed by multiple regression (least square estimation) with SMR as dependent variable and the other social factors as independent variables. METHODS All 192 district health authorities in England and Wales. RESULTS There was a significant association between UPA-score and SMR (R-square = 0.53, p = 0.0001). The second model included the four variables: % unskilled, % unemployed of economically active population, % of residents living in overcrowded households, and % of population from New Commonwealth and Pakistan. This model explained about 77% of the variance. All variables were significant. All coefficients except ethnicity were positive, i.e. the higher the coefficients, the higher the SMR. CONCLUSIONS This paper shows that social deprivation, unemployment and overcrowding were related to mortality in district health authorities in England and Wales. The finding that ethnicity was inversely related to general mortality might be important and needs to be further analysed.