Estradiol-17β (E2), administered systemically to rhesus monkeys during the follicular phase of the menstrual cycle, induces atretic changes in the microenvironment of the dominant follicle (DF), which results in its demise. It has been proposed that this effect of E2 represents a direct action at the ovarian level. The present study was designed to test this hypothesis, using local treatment with E2. After identification of the DF during laparoscopy on day 6 of the cycle, female monkeys were laparotomized and their ovaries exposed. Either corn oil (20 μl, controls) or E2 (100 μg ) in oil vehicle (experimentals) was injected into the ovary near the DF. In control animals, preovulatory release of gonadotropins and ovulation were normal in five of six animals, with cycle and luteal phase lengths of 27.8 ± 2.2 days and 14.6 ± 2.5 days, respectively (x̄ ± S.D.). Conversely, in only one of six animals in the experimental group did ovulation occur at the expected time (P < 0.05). In the other five treated animals, E2 induced atresia of the DF and significantly extended cycles (35.4 ± 5.4 days) without affecting luteal phase lengths (12.0 ± 1.4 days). Concentrations of estrogen in peripheral sera of some animals were increased transiently at 6 h after injection of E2 but returned to normal by 12 h; this duration of estrogen exposure is far less than the 24 h required to induce atresia of the DF in previous studies. At 6 h after injection of E2, there was a statistical difference between controls and experimentals in concentrations of circulating estrogen; however, these changes were apparently not enough to alter pituitary secretion of follicle-stimulating hormone or luteinizing hormone. These data support the hypothesis that E2 can induce atresia of the DF in rhesus monkeys by acting locally at the ovary.