The present study was concerned with the control of luteal activity in female rats which had been treated neonatally with 1.25 mg testosterone propionate (TP). Treatment of such rats in adulthood with 15 i.u. human chorionic gonadotrophin induced ovulation followed by a period of luteal activity. The two daily surges of prolactin secretion, typical for a period of luteal activity in the normal female rat, were not observed in TP-treated females. Instead, higher basal levels of prolactin were observed in TP-treated females than in normal female rats. Furthermore, uterine traumatization at 5 days after ovulation did not result in the formation of decidual tissue. In intact TP-treated females luteal activity, induced and temporarily sustained by an ectopic pituitary transplant, persisted after removal of the pituitary graft. In contrast, in TP-treated females which had been ovariectomized on day 25 of age and had received an ovarian transplant before induction of the luteal phase, luteal activity ended within a week after removal of the ectopic pituitary gland. Females treated with TP which had been ovariectomized on day 25 of life had lower plasma levels of prolactin and higher levels of dopamine in hypophysial stalk plasma than intact TP-treated females when measured at 4 months of age. Treatment of ovariectomized rats with oestradiol-17 beta increased levels of prolactin in plasma and lowered levels of dopamine in hypophysial stalk plasma. It is concluded that the control of luteal activity in TP-treated females shows 'male' characteristics. However, the presence of the ovaries in such rats leads to decreased hypothalamic release of dopamine and increased plasma levels of prolactin, probably due to increased oestrogen levels. These increased levels of prolactin are sufficient to maintain luteal activity.