To determine correlates of and recent trends in aspirin use in middle-age men and women, we analyzed data from population-based samples selected in four U.S. communities. Aspirin use (during a 2-week period preceding the study examination) was more prevalent in whites than in blacks (30% vs 11%; p < 0.001) and in men than in women among whites (31% vs 28%; p < 0.002) but not blacks (10% in both sexes). In all four race and sex groups, there was a graded positive relation between estimated coronary heart disease (CHD) risk and age-adjusted prevalence of aspirin use. For example, 33% of CHD-free white men who reported diagnoses of hypercholesterolemia and hypertension and had ever smoked reported aspirin use as compared with 25% of their risk factor-free counterparts (p < 0.001). Among men with symptomatic CHD or at high risk for CHD, aspirin use increased by four percentage points between 1987 and 1989 in conjunction with the publication of results from the aspirin primary prevention trials. However, nearly 50% of participants reporting a history of myocardial infarction apparently did not take aspirin regularly.