Although testosterone supports all phases of spermatogenesis in primates, FSH is obligatory for quantitatively normal spermatogenesis. To further investigate the mechanism of action of FSH on spermatogenesis, eight adult male rhesus macaques were hypophysectomized and supplemented daily with cortisone acetate (5 mg/kg BW, sc) and T4 (50 mg/animal, orally). Complete pituitary ablation was established by 1) a decline in mean testicular volume to 8% of the prehypophysectomy value; 2) a failure to secrete gonadotropins in response to 50 micrograms GnRH, iv; and 3) an absence of pituitary in the sella turcica on postmortem examination. Testosterone-filled SILASTIC brand capsules (Dow Corning) were implanted sc to restore testicular testosterone to normal levels. Once the testes had achieved maximum growth under testosterone stimulation alone, the animals were implanted with indwelling venous catheters. In four animals, a pulsatile infusion of human FSH (one pulse of 4 IU/kg BW every 3 h) was administered for 12 days, and the other four monkeys received vehicle. Testosterone replacement continued throughout the experiment. At the termination of the 12 days of FSH stimulation or vehicle administration, the right testes were removed. The left testes were removed 22 days later to investigate whether testosterone was able to maintain the effects, if any, of FSH stimulation. Portions of each testis were fixed in Bouin's solution and subsequently prepared for histological examination, whereas other portions were frozen in liquid nitrogen for determination of testicular testosterone content. Five hundred cross-sections of seminiferous tubules in periodic acid-Schiff-hematoxylin-stained histological sections were randomly selected from each testis. The stage of the seminiferous epithelial cycle in these sections was identified, and the germ cells and Sertoli cells were counted in each. All cell counts were corrected by the Abercrombie method and expressed per cross-section of seminiferous tubule. Treatment with FSH for 12 days failed to influence the numbers of either Sertoli cells or Ad and Ap stem spermatogonia. In striking contrast, the number of all four generations of differentiated (B1, B2, B3, and B4) spermatogonia were significantly amplified by stimulation with human FSH for 12 days. As reflected by the analysis of the left testes collected 22 days after termination of the gonadotropin treatment, the progeny of these B spermatogonia were not maintained in the absence of FSH. In conclusion, the results of the present study indicate that the action of FSH to quantitatively maintain spermatogenesis in the rhesus monkey is mediated by a selective amplification of B spermatogonia.