A composite measure of subclinical vascular disease has been developed in the Cardiovascular Health Study (CHS). In previous reports, we measured the prevalence of subclinical disease among the original 5201 participants in the CHS, the relationship of risk factors to subclinical disease, and the association of subclinical disease to clinical coronary heart disease. In 1992 to 1993 (year 4 of the study), a larger cohort of 424 black women and 248 black men was added to the study. In this study, we have compared the prevalence of subclinical disease among blacks and whites in the CHS and the association with cardiovascular risk factors. The prevalence of subclinical disease for all participants (aged > or =65 years) was 41.3% for white women, 39.7% for black women, 41.9% for white men, and 43.7% for black men. The prevalence increased with age. The risk factor associations for subclinical disease were similar among blacks and whites. In multivariate analysis, age, systolic blood pressure, LDL cholesterol, smoking, and family history of myocardial infarction were independently associated with subclinical disease among both black and white women, while for white men, systolic blood pressure, use of antihypertensive medication, smoking, body mass index, and diastolic blood pressure (inverse) were related to subclinical disease. In black men, blood triglyceride level, use of antihypertensive medications, and family history of myocardial infarction (inverse) were associated with subclinical disease.