To evaluate the physiology of inhibin B in the human male, we measured serum concentrations in normal adult men and men with isolated GnRH deficiency before and during long-term replacement with pulsatile GnRH. At baseline, inhibin B levels in the GnRH-deficient men (n = 31) were significantly lower than normal controls (85 +/- 10 pg/mL vs. 239 +/- 14 pg/mL; P < .01) and correlated positively with pretreatment testicular volume (r = .80, P = .001) and a history of spontaneous puberty, suggesting additional maturational influences on the both testicular volume and inhibin B secretion. Pulsatile GnRH administration was associated with significant increases in inhibin B, with levels averaging 108 +/- 7 pg/mL when serum LH, FSH, and T concentrations had reached the normal adult male range (n = 22; P = .02 vs. baseline). Continued GnRH administration for at least an additional year was not associated with further increases in inhibin B concentrations. Throughout the course of long-term pulsatile GnRH replacement, serum FSH levels were negatively correlated with inhibin B concentrations (e.g. r = -.71, P < 0.01; n = 14 treated 12 months after normalization of T). Although inhibin B concentrations did not correlated with sperm density during therapy, rates of fertility were higher in patients with higher baseline levels (inhibin B > or = 60 pg/mL). Increases in serum concentrations of inhibin B occurring during GnRH replacement demonstrate the gonadotropin regulation of gonadal inhibin B secretion. However, the variation in baseline inhibin B levels before GnRH administration suggests an additional gonadotropin-independent level of modulation. The negative correlation between FSH and inhibin B secretion in GnRH-deficient men receiving long-term GnRH replacement is consistent with a putative role of inhibin B in the negative feedback regulation of FSH, although direct confirmation of this role requires further investigation.