This study was undertaken to clarify the complex relationship between poverty and race with disease-specific mortality. Data from the 1987 National Health Interview Survey and the 1986 National Mortality Followback Survey were used to estimate standardized mortality ratios (SMRs) for various categories (all causes, all cancers, noncancerous medical causes, lung and breast cancers, ischemic heart disease, and cerebrovascular disease) associated with income below the poverty line and were compared with those with adequate or better than adequate income. All SMRs were substantially elevated. The SMRs were not appreciably affected by adjustments for confounding by alcohol consumption, occupation, or smoking. Sex-specific SMRs of blacks relative to whites with the exception of ischemic heart disease were significantly elevated for males but not for females with the exception of the SMR for all causes. However, when adjusted for per capita income within the family, black mortality never significantly exceeded that of whites. These results strongly support the conclusion that mortality differences between blacks and whites are due to differences in income and related factors. However, it is unclear why poverty should be associated with elevation in cancer relative risk.