The detection of pulsatile ovine LH (oLH) secretion in the sheep fetus by 81 days gestation (term 147 days), the suppression of fetal gonadotropin secretion by chronic administration of an LH-releasing factor agonist or antagonist, and the capacity of N-methyl; D-aspartate (a neuroexcitatory amino acid analog) to evoke a fetal oLH pulse strongly support a functional LH-releasing factor pulse-generator in the ovine fetus. In light of the sex difference in fetal gonadal function and gonadotropin secretion before day 114, we postulated that fetal castration would have a discordant effect on the pattern of gonadotropin secretion in males and females. Fetal sheep were either castrated (male = 11; female = 9) or sham operated (male = 9; female = 6) at 110-115 days gestation. Chronic indwelling arterial and venous catheters were implanted, and animals were studied for up to 30 days. During each study period fetal arterial blood samples were drawn every 15 min for 5 h and the plasma assayed for oFSH and oLH by specific RIAs. Multiple studies were performed on each fetus. In all fetuses (both intact and castrated) a decrease in oLH pulse frequency occurred after day 130. In female fetuses before day 130, castration had no effect on mean oLH pulse frequency (sham, 0.72 +/- 0.19 pulses/5 h; castrate, 0.50 +/- 0.13 pulses/5 h; P > 0.05). After day 130, pulsatile oLH secretion decreased in both intact and castrated female fetuses to undetectable levels during the sampling period. In contrast, castration significantly (P < 0.001) increased mean oLH pulsatility in males before and after day 130 (< 130 days, sham, 1.06 +/- 0.24 pulse/5 h; castrate, 2.70 +/- 0.22 pulse/5 h; > 130 days, sham, 0.18 +/- 0.12 pulses/5 h; castrate, 1.65 +/- 0.26 pulses/5 h). Mean oLH pulse amplitude was increased by castration only in the male fetuses (sham, 3.89 +/- 0.87 ng/ml; castrate, 6.02 +/- 0.39 ng/ml; P < 0.05). oFSH pulses were infrequent in both sexes and not influenced by castration. The mean plasma concentration of oFSH was greater in intact female fetuses than in intact males (female, 5.65 +/- 1.15 ng/ml vs. male, 2.07 +/- 0.45 ng/ml; P < 0.01). Castration increased the mean value for plasma oFSH in males (4.40 +/- 0.43 ng/ml; P < 0.001) but had no effect in females (3.83 +/- 0.64 ng/ml; P > 0.05). The striking sex difference in the effect of fetal castration on pulsatile oLH secretion and mean plasma oFSH provide evidence for a fundamental organizational effect of fetal testicular hormones, especially testosterone, on the development of the fetal hypothalamic-pituitary gonadotropin unit.