The well-known antioxidant properties of selenium have been linked to a lower incidence of cardiovascular disease in humans, but the findings remain controversial. To explore whether the plasma selenium level predicts risk of myocardial infarction (MI), we analyzed prospectively collected plasma samples in a nested case-control study among participants in the Physicians' Health Study, a randomized trial of aspirin and beta-carotene. Blood specimens were collected between August 1982 and December 1984 and stored at -80 degrees C. All infarcts were documented by medical records. This study is based on 251 subjects who had infarctions and an equal number of healthy controls, matched by age, smoking status, and time from randomization. The mean +/- SD levels of plasma selenium were 114.4 +/- 15.1 ng/g in the cases with MI, and 113.2 +/- 15.7 ng/g in controls (paired t = 0.94, p = 0.35). Conditional logistic regression analysis by quintile of plasma selenium levels showed no suggestion of any protective effect of selenium; subjects in the highest quintile had a relative risk of 1.27 (95% confidence interval 0.71 to 2.29) when compared with the bottom quintile, and 1.53 (95% confidence interval 0.61 to 3.84) after adjustment for other cardiovascular risk factors. These data provide no evidence for an association between increased plasma selenium and reduced risk of MI at the current levels of selenium intake within the U.S.