A pedigree of complex structure, comprising 195 individuals, is shown to be segregating for an autosomal dominant gene for hypercholesterolemia. The same pedigree shows two groups of individuals with respect to plasma triglyceride levels; the cause of this appears to be independent of the locus demonstrated, which accounts for over 50% of the variability in cholesterol levels in the kindred studied. The power of bivariate analyses of multigenerational data, as used in this study, is discussed. The lipid disorders in this family, however, differ from the usual familial hypercholesterolemia described by Harlan et al. (1966) in that there is a group of individuals who have elevated triglycerides, there is a substantial amount of premature cardiovascular disease, and not very many individuals have tendonous xanthomas. It is of course possible that in this family the genetic entity is a variant of the usual hypercholesterolemia mutant; or it may be the usual mutant, acting in concert with some other unknown factor. The cause of the two groups with respect to triglycerides is not clear; but, whether it is genetic or environmental, it is independent of the segregation for hypercholesterolemia.