In this study we examined the independent and interactive effects of lifetime patterns of drinking and smoking on cognitive performance in elderly African Americans. A sample of 230 individuals with varying histories of alcohol and cigarette use was drawn from the Hillsborough Elder African American Life Study, a community-based, cross-sectional study of older adults aged 60 to 84. Dependent variables were the results of a neuropsychological battery that provided measures of general cognitive ability, executive function, and memory. Specifically, our study addressed (1) whether individuals with a lifetime history of sustained smoking and/or drinking show lower levels of cognitive performance in comparison to lifetime abstainers, (2) whether cumulative lifetime doses of alcohol or cigarettes, or of the two substances in interaction, have an effect on cognition, and (3) whether individuals who have histories of periodic, intense use of either alcohol or cigarettes show lower levels of cognitive performance in comparison to lifetime abstainers. When significant results were obtained, effect sizes were small, not exceeding 5% of the variance. A single exception occurred for the intensity analyses, in which drinking explained approximately 16% of the variance in global cognitive ability after adjusting for the contributions of control variables. In these analyses, drinking was found to have a U-shaped effect on global cognitive ability and total acquisition in the memory trials. Specifically, moderate users performed at a lower level than abstainers or heavy users, who did not differ from each other. (JINS, 2003, 9, 690–697.)