Data from a community-based sample of 2,812 men and women aged 65 years and over, living in New Haven, Connecticut, were used to examine the associations between blood pressure, smoking, diabetes, anginal chest pain, and relative weight and 6-year incidence (1982-1988) of myocardial infarction and coronary heart disease mortality. Multivariate logistic regression analyses revealed that history of diabetes was associated with increased risk of incident myocardial infarction among women (odds ratio (OR) = 3.20; 95% confidence interval 1.46-7.01), while higher relative weight was a significant risk factor among men (OR = 3.46, 95% confidence interval 1.34-8.95 for middle vs. lowest tertile and OR = 3.24; 95% confidence interval 1.10-9.53 for highest versus lowest tertile). For coronary heart disease mortality, age and diabetes (OR = 4.47, 95% confidence interval 1.85-10.79) were associated with increased risk among women, as was current smoking (OR = 3.96; 95% confidence interval 1.66-9.45). Among men, age, prevalent heart disease, and use of antihypertensive medication (OR = 1.84; 95% confidence interval 1.13-3.00) were risk factors for coronary heart disease mortality. These risk estimates and the relatively high prevalences of these risk factors suggest that attributable risks may be substantial. Observed sex differences as well as differences in patterns of risk factor associations for the two endpoints suggest that there may be different risk profiles for older men and women and for different coronary heart disease endpoints.