Genetic variability has been implicated as a significant contributor to the variation in levels of lipids, lipoproteins, and apolipoproteins (apos) through a variety of direct and indirect investigations. Among the direct investigations, apo E has been shown to be polymorphic and to explain a small but statistically significant proportion of the variability in cholesterol. The apo E polymorphism was typed in 964 randomly selected Mexican-Americans from Starr County, Tex., and its effects determined on levels of cholesterol, triglycerides, total high density lipoprotein (HDL) cholesterol, subfractions (HDL2 and HDL3), alpha- and beta-lipoprotein cholesterol, low density lipoprotein (LDL) cholesterol, and apos A-I, A-II, B, C-II, C-III, and E. Effects are reported for the entire sample and in each of three groups, namely, premenopausal females, postmenopausal women, and males. In the entire sample, significant effects were observed on cholesterol, beta-lipoprotein cholesterol, LDL, apo B, and apo E. There is evidence for significant physiological interaction of the apo E polymorphism effect in females by menopausal status. This is most evident for apo E levels, in which 5.9% of the variability in the entire sample is explained by the apo E polymorphism. In premenopausal females, however, the polymorphism accounts for 27.5% of the variability. In postmenopausal women and males, there is no significant effect. It is shown that the apo E polymorphism can be treated as a two-locus, two-allele system. Doing so identifies substitutions in amino acid position 158 as the mediators of most of the observed effects of this polymorphism.