Background. Studies of underlying differences in adult mortality between black and white individuals in the USA have been constrained by limitations of data or small study size. We investigated the extent to which differences in socioeconomic position between black and white men contribute to differences in all-cause and cause-specific mortality. Methods. 361,662 men were screened for the Multiple Risk Factor Intervention Trial between 1973 and 1975, in 22 sites. Median family income of households by zipcode (postal) area of residence was available for 20,224 black and 300,685 white men as well as data on age, cigarette smoking, blood pressure, serum cholesterol, previous heart attack, and treatment for diabetes. We classified deaths during 16 years of follow-up into specific causes and compared differences in death rates between black men and white men, before and after adjustment for differences in income and other risk factors. Findings. Age-adjusted relative risk of death (black vs white) was 1.47 (95% CI 1.42-1.53). Adjustment for diastolic blood pressure, serum cholesterol, cigarette smoking, medication for diabetes, and previous admission to hospital for heart attach decreased the relative risk to 1.40 (1.35-1.46). Adjustment for income but not the other risk factors decreased the risk to 1.19 (1.14-1.24) and adjustment for other risk factors did not alter this estimate. For cardiovascular death, relative risk on adjustment for income was income was decreased from 1.36 to 1.09; for cancer from 1.47 to 1.25; and for non-cardiovascular and non-cancer deaths from 1.71 to 1.26. For some specific causes of death, including prostate cancer, myeloma, and hypertensive heart disease, the higher death rates among black men did not seem to reflect differences in income. Rates of death for suicide and melanoma were lower among black than white men, as were those for coronary heart disease after adjustment for income. Interpretation. Socioeconomic position is the major contributor to differences in death rates between black and while men. Differentials in mortality from some specific causes do not simply relect differences in income, however, and more detailed investigations are needed of how differences are influenced by environmental exposures, lifetime socioeconomic conditions, lifestyle, racism, and other sociocultural and biological factors.