It has been suggested that growth hormone (GH) may play a regulatory role in male reproductive function. To express full anabolic effect in immature boys testosterone apparently requires the presence of GH. In GH-deficient adults, GH replacement therapy exerts a variety of anabolic actions, some of which are similar to the effects of gonadal steroids. However, little is known about the potential effects of GH on gonadal steroids and on dynamic tests of pituitary-gonadal function in adults with GH deficiency. We evaluated the pituitary-gonadal axis in a 4-month double-blind, placebo-controlled GH study in 13 young males with childhood-onset GH deficiency of which 6 had isolated GH deficiency. GH treatment significantly increased serum levels of total IGF-I from 98 (68) to 323 (126) microg/l, free IGF-I from 0.48 (0.47) to 2.24 (1.66) microg/l, IGFBP-3 from 1,874 (1,178) to 3,520 (778) microg/l and ALS levels from 9,182 (5,524) to 16,872 (6,278) microg/l (all p <0.0001). We found no differences in basal testosterone levels in the 13 patients between the GH and placebo treatment periods (21.9 (5.1) vs. 24.5 (8.1) nmol/l, nonsignificant). Furthermore, no effect of GH on the testicular response to hCG after 72 h was seen compared to placebo (36.2 (6.4) vs. 38.8 (10.3) nmol/l). In addition, no differences existed in basal SHBG, DHT, free testosterone, delta4-adion and DHEA-S levels. There were no statistically significant differences in maximal FSH and LH response to a GnRH challenge between the GH and the placebo periods (15.7 (5.3) vs. 18.0 (8.8) U/l and 47.0 (26.4) vs. 40.4 (26.5) U/l, respectively). Furthermore, there was no effect on cortisol responses after ACTH between the GH and the placebo periods. However, significantly higher estradiol levels were seen after GH treatment (110 (50) pmol/l) compared to after placebo (89 (34) pmol/l, p = 0.03). Prostate-specific antigen levels decreased after GH treatment compared to after placebo (0.42 (0.54) vs. 0.47 (0.48) microg/l) and this difference almost reached statistical significance (p = 0.059). Inhibin-B levels were significantly lower in hypogonadal patients substituted with androgens, but GH had no effect on inhibin-B levels. In conclusion, GH replacement therapy in 13 GH-deficient young adult males resulted in significant increases in total and free IGF-I as well as in ALS levels in all patients, but had no significant effect on: (1) the pituitary FSH and LH response to GnRH; (2) basal and hCG-stimulated levels of androgens and SHBG; (3) basal inhibin-B levels; (4) ACTH-stimulated cortisol secretion. By contrast, GH administration had subtle anti-androgenic effects in terms of elevated elevated estradiol levels and decreased prostate-specific antigen levels, although both parameters remained within the normal range. Thus, at the level of blood chemistry the effects of GH administration do not appear to involve major alterations in the pituitary-gonadal axis.