Although sex hormones appear to be importantly involved in the development of coronary heart disease, apparently no study has yet reported an alteration in an endogenous sex hormone level in relation to coronary heart disease in women. In an attempt to determine whether any sex hormone abnormality might be a factor in the development of myocardial infarction in women, estradiol and testosterone, as well as sex hormone-binding globulin, insulin, dehydroepiandrosterone sulfate, and risk factors for myocardial infarction, were measured in relation to the degree of coronary artery disease (CAD) in 60 postmenopausal women undergoing coronary angiography. In a multiple-regression analysis with the degree of CAD as the dependent variable and free testosterone (FT), estradiol, age, body mass index, systolic blood pressure, cholesterol, smoking, and insulin as independent variables in the model, only FT (P < .008) and cholesterol (P = .01) were significantly related to the degree of CAD, both positively. To exclude a possible confounding effect due to prior myocardial infarction, the multiple-regression analysis was repeated for the subgroup of 49 patients remaining after excluding the 11 patients who had ever had a myocardial infarction; again only FT (P < .04) and cholesterol (P = .05) were significantly related to the degree of CAD. Neither total testosterone in place of FT nor HDL cholesterol in place of total cholesterol in the model was significantly related to CAD. Sex hormone-binding globulin and dehydroepiandrosterone sulfate, added individually to the model, showed no significant relationship to CAD. These results raise the possibility that in women an elevated FT level may be a risk factor for coronary atherosclerosis.