Summary. Rats were given s.c. implants of high (HT) or low (LT) doses of testosterone and 10 days later hypophysectomy or sham-operation was performed. The rats were killed after 50 days. Unilateral efferent duct ligation was performed 16 h before death to measure seminiferous tubule fluid production and the increment in testicular inhibin values (inhibin production). Inhibin levels in testis cytosols were measured by a pituitary cell culture bioassay. The LT implants maintained serum testosterone at control values and decreased testicular weight whereas HT implants raised serum testosterone 3-fold and maintained testicular weight at 75–85% of pretreatment levels. In intact rats, LT implants caused no change in testicular inhibin content but decreased inhibin production; no significant changes occurred with HT implants. After hypophysectomy both values were significantly suppressed and could not be maintained by HT or LT implants. However, the HT implants partly restored inhibin production despite their inability to influence testicular inhibin content. In contrast, tubule fluid production depended mainly on intratesticular testosterone levels and occurred normally in intact or hypophysectomized rats with HT but not LT implants. These results indicate that inhibin and seminiferous tubule fluid production, both functions of the Sertoli cell, are under different hormonal control. The maintenance of inhibin production by the testis requires the support of pituitary hormones, presumably FSH, while seminiferous tubule fluid production requires testosterone, presumably through LH stimulation of Leydig cells. These findings are consistent with the hypothesis that inhibin is produced in response to trophic stimulation by FSH.