Estrogen (E) reduces bioassayable GH-dependent serum somatomedin (SM) activity in acromegalics without affecting plasma growth hormone (GH) levels and inhibits the rise of SM activity normally produced by GH administration in GH-deficient subjects. We have now investigated the effect of E administration on serum SM activity and on plasma GH and prolactin (PRL) in 6 adult male subjects without pituitary pathology. Chronic E administration (ethinyl estradiol 0.5 mg/day for 7 to 70 days) reduced serum SM activity by 40 to 62% in each of 4 subjects (P less than 0.02 to less than 0.001). In 3 of the subjects, basal GH levels increased by 75 to 300% (P less than 0.05 to less than 0.001) and basal PRL levels increased by 90 to 200% (P less than 0.01 to less than 0.001). While iv administration of normal saline did not significantly affect either SM or GH, iv administration of E (bolus injection of 25 mg conjugated estrogens, USP) to 5 subjects resulted in: a) a 46 to 80% decrease in serum SM activity in all subjects, proceeding with an apparent half-life of 2 hours, becoming significant (P less than 0.05) at 2 hours (1 subject) to 3 hours (4 subjects), maximal at 6 hours, and persisting for 12 to 24 hours; b) GH elevation to 3 to 16 times baseline level (P less than 0.01) at 2 to 3 hours in 4 subjects; and c) no significant change of PRL levels in any subject. The mean GH response to iv E was maximal at a time (2 hours) when the mean SM activity had decreased only 20% and subsided well before the nadir of SM activity. The one patient without GH response to chronic or acute E administration may have been affected by absorption of triamcinolone being applied topically during the study. These results demonstrate that in males with normal pituitary function, E reduces serum SM activity, enhances basal GH and PRL secretion, and, upon iv injection, stimulates acute GH release. Although opposite chronic E effects upon GH and SM activity support a putative negative SM-GH feed-back mechanism, iv E administration apparently provokes acute GH release by a different mechanism. The half-life of serum SM activity in the human is probably much shorter than previously estimated.