We investigated the mediating role of genetic factors in the clustering of hypertension, diabetes, and obesity, using the twin registry maintained by the National Academy of Sciences--National Research Council. The study sample included 2,508 male twin pairs born between 1917 and 1927 who responded to a mailed questionnaire that covered demographic variables, cardiovascular risk factors, and health behaviors. The incidence of hypertension and diabetes in this cohort was ascertained from subjects' self-report of a physician diagnosis and/or the use of prescription medications. The body-mass index calculated from self-reports of height and weight was used as a measure of obesity. Descriptive analyses indicated probandwise concordance rates of 34.0%, 31.2%, and 32.7%, respectively, for the joint occurrences of hypertension and diabetes, hypertension and obesity, and diabetes and obesity in MZ twin pairs. Corresponding concordance rates in DZ twin pairs were 8.1%, 14.9%, and 2.8%. The probandwise concordance for the clustering of all three conditions in the same individuals was 31.6% in MZ pairs and 6.3% in DZ pairs (relative risk 5.0; X2(1) = 2.6; P < .15). Multivariate genetic modeling of the correlation in liabilities to develop these conditions suggested the presence of a common latent factor mediating the clustering of hypertension, diabetes, and obesity in this twin sample. This common factor was influenced by both genetic and environmental effects (59% 8 genetic, 41% environmental). The genetic influences on the common latent factor were due to dominant rather than additive sources; the environmental influences appeared to be specific rather than shared by co-twins.