The cardiovascular protective effects of estrogen are known to be mediated by its beneficial effects on lipid metabolism and its direct actions on the vessel wall. The latter can be mediated by a specific receptor for estrogen present on smooth muscle cells and endothelial cells. The gene for the receptor (the classic estrogen receptor [ER]) has three known polymorphisms, Pvu II, Xba I, and B-variant polymorphisms, which are reportedly associated with receptor expression and altered receptor function and with some disorders including breast cancer, hypertension, and spontaneous abortion. However, the significance of genetic variations of the ER in vascular diseases has not been reported. We have examined the association between coronary artery disease (CAD) and the three polymorphisms in ER. Genotypes (P1/P2, X1/X2, and B-wild type/B-variant type) were determined in 87 men and postmenopausal women with myocardial infarction or angina pectoris whose lesions were confirmed by coronary angiography, as well as from 94 control individuals from the general population with no coronary heart disease and normal resting ECG. For B-variant polymorphism, all individuals examined had B-wild type, which contrasts with the reported allele frequency for B-variant type (0.1) in the white population. Genotype distributions and allele frequencies of Pvu II or Xba I polymorphisms were not significantly different between control subjects and patients (P > .05 for Pvu II or Xba I genotypes; P > .05 for Pvu II or Xba I allele frequencies). When the allele frequencies were analyzed separately by sex, there was still no statistically significant difference for both polymorphisms (P > .05 for men; P > .05 for women). No association was found between the polymorphisms and the angiographic severity of CAD. Total cholesterol, triglyceride, or HDL-cholesterol levels were not significantly different among ER genotypes. These findings suggest that the three polymorphisms in ER are not associated with the prevalence and severity of CAD and that the polymorphisms are unrelated to the serum lipid levels in control subjects and patients.