Digoxin has been reported to induce feminizing effects in man. It does not compete for estradiol cytosol receptors in human breast carcinoma cells, however, and has no uterotrophic effect. We therefore investigated whether feminization might be due to digoxin action on plasma concentrations of sex steroids. Six healthy men (31.5 +/- 4 yr old) received therapeutic doses of digoxin for 43 days. We measured plasma concentrations of testosterone, androstenedione, dehydroepiandrosterone, estrone, estradiol, progesterone, 17-hydroxyprogesterone, cortisol, and aldosterone. During 35 days on digoxin levels of these steroids remained in the normal range and there was no change from before-drug values. Digoxin was in the therapeutic range of 1.9 +/- 3 nmol/l throughout. After stimulation by adrenocorticotropic hormone or human choriongonadotropin, the rise in plasma steroids was in the same range as when digoxin was given, as well as 16 wk after it had been discontinued. A normal rise in luteinizing hormone after luteinizing hormone-releasing hormone showed that the hypothalamogonadal feedback was not altered by digoxin. Free testosterone, estradiol, and cortisol concentrations under basal conditions and after stimulation were also the same before and after drug. It is concluded that the estrogen-like activity of digoxin cannot be explained by altered steroid availability from plasma. Feminizing effects attributed to digoxin may be caused by other conditions known to influence sex steroid hormones that are common in patients with heart disease. Our data suggest that digoxin may be the preferred digitalis therapy to avoid feminizing effects.