To probe the relationship between the size of the Sertoli cell population, established during perinatal development, and production of germ cells in the adult testis, a Sertoli cell-depleted rat model was developed. This was accomplished by delivering an antimitotic drug, cytosine arabinoside (araC), directly to the testis of newborn pups. Initial studies of these araC-treated neonates indicated that 1) the drug is cleared rapidly from the testis; 2) it substantially reduces the level of Sertoli cell proliferation; 3) Sertoli cell division ceases at a normal time in spite of the previous drug treatment; and 4) araC itself has no residual effect on germ cell proliferation, which begins several days after the injection. Pups given araC were allowed to reach maturity, and their testes were perfuse-fixed for light microscopic morphometry. When the numbers of Sertoli cells in adult rats given araC as were compared with those in normal littermates, a 54% decrease in the size of the Sertoli cell population was detected in treated rats, now referred to as Sertoli cell-depleted. Moreover, when round spermatids were quantified and compared in normal and Sertoli cell-depleted adults, testes of the latter were found to contain 55% fewer round spermatids. Since, in the araC-treated group, the decrease in Sertoli cell population size was paralleled by a reduction in spermatid production of equal magnitude, the number of round spermatids per Sertoli cell was essentially identical in normal and Sertoli cell-depleted animals. Measurements of serum androgen-binding protein (ABP) and FSH in both groups indicated that the circulating level of ABP in Sertoli cell-depleted rats was approximately half, and the concentration of FSH approximately twice, that in normal animals. Thus, even though FSH is elevated in Sertoli cell-depleted rats, the production of ABP per Sertoli cell is unchanged. In addition, collective volume of Leydig cells and ventral prostate weights were normal in the Sertoli cell-depleted group, suggesting that Leydig cell function in these rats is normal. In summary, a Sertoli cell-depleted rat model has been produced by interfering specifically with Sertoli cell proliferation early in postnatal life, before onset of germ cell division. Moreover, our findings with this model indicate that production of normal numbers of germ cells in adults depends, at least in part, on the size of the Sertoli cell population. Thus, our observations identify the perinatal period, when the Sertoli cell population is established, as critical for development of quantitatively normal spermatogenesis in the adult.