[en] Standardization of mortality for variation in sociodemographic characteristics as well as for age, sex, and race is required for improved detection of changes in mortality around point sources of pollution. Mortality by level and selected causes was studied in 191 counties within 80 km of 15 power generation sites chosen to represent the diversity of US socioeconomic, demographic, and geographic features. Mortality data for 1969 to 1971 were evaluated against selected 1970 US census, air pollution, climate, and cigarette and beer consumption data. A mortality process scheme was developed and used to rationalize the selection of independent variables. Applying this technique to regressions of mortality on air pollution and social-demographic variables, only female mortality was found to have strong independent or interactive effects with air pollution. For cause-specific mortality, the explained variance was small for diseases with major genetic components and larger for diseases with strong socioeconomic components. Testing of the regression equations indicated that standardization for socioeconomic factors may be a useful strategy on a regional basis