The aim of this study was to investigate the progression of germ cell populations through the rat spermatogenic cycle when spermatogenesis was suppressed by LH withdrawal through the use of a combination of testosterone (T) and estradiol (E) and then reinitiated by the administration of high doses of T. Adult Sprague-Dawley rats received 3-cm T and 0.4-cm E silastic implants for 6, 8, or 12 wk to suppress spermatogenesis followed by high- dose T (24-cm implants) for up to 12 wk to reinitiate spermatogenesis. The number of spermatogonia, primary spermatocytes, and round spermatids per testis was established by stereological techniques, and the elongated spermatid number was determined by the testicular content of nuclei resistant to homogenization in Triton X-100. Suppression for 6-12 wk resulted in moderate and significant (p <0.05) reductions in the numbers of type A and type B spermatogonia (to 44-59% of control levels), preleptotene (68-72% of control), and leptotene/zygotene spermatocytes (62-79% of control) as well as in the numbers of stage I-VII (56-69% of control) and stage VIII-XIV (35-43% of control) pachytene spermatocytes. Round spermatids were suppressed to 29- 45% of control levels (p <0.05) while elongated spermatids were undetectable. The hourly production rates of germ cells (calculated using published time divisors) were used to study the cellular conversions through spermatogenesis (based on the ratios of the hourly production rates) and revealed that T withdrawal consistently abolished the conversion of round to elongated spermatids. The duration of suppression (6, 8, or 12 wk) had no effect on the degree to which germ cell populations or conversions were reduced. In response to high-dose T administration, spermatogonial and spermatocyte numbers and production rates (up to stage I-VII pachytene) remained suppressed, while stage VIII-XIV pachytene spermatocytes showed an increase of borderline significance. On the other hand, round and elongated spermatid numbers and production rates increased significantly (to 81% and 78% of control, respectively) and their conversion was normalized, i.e., the spermiogenic process was restored to a level consistent with the numbers of earlier germ cells proceeding through the cycle. These data suggest that, in the presence of low T levels, spermatogenesis proceeds at ~65% of normal levels between the spermatogonial and round spermatid stages, irrespective of the duration of T-induced suppression. This is followed by a precipitous decline in elongated spermatid number that is attributed to the disappearance of round spermatids. The normalization of maturation from round to elongated spermatids is the predominant action of T in the restoration of spermatogenesis. The failure of high-dose T to quantitatively restore spermatogonial number (and consequently other germ cell populations and testicular weight) suggests that a higher T dose and/or other factors (e.g., Leydig cell products) are needed for normal spermatogonial recruitment.