Anterior pituitary gland fragments were obtained from female Wistar-derived rats on dioestrus or pro-oestrus and perifused in Biogel columns in vitro. They were subjected at the beginning of each of 5 h of perifusion to a volley of 6 1-min pulses of luteinising hormone-releasing hormone (LHRH, 10 nM), given 4 min apart. Luteinising hormone (LH) was measured by radioimmunoassay in sequential 2-min fractions of the perifusate. Pituitary glands removed at 14.00 h on dioestrus showed a characteristic pattern of sensitisation followed by desensitisation to the repeated volleys of LHRH, whereas tissues removed at 10.00 or 14.00 h on pro-oestrus showed no evidence of desensitisation over the 5-hour period, the response to each LHRH volley being greater than the preceding one. Estradiol (E, 3–100 pg/ml) added to the medium from the start of perifusion had no significant effect on the pattern of response from tissues removed on pro-oestrus, but the highest concentration significantly enhanced the response of dioestrus pituitaries to all but the last of the LHRH volleys. Progesterone (P, 1–50 ng/ml) added to the medium produced a dose-related inhibition of the response of pro-oes-trous tissues to the LHRH volleys. Groups of animals were ovariectomised (OVX) on dioestrus and used for experiment the next morning. OVX at 10.00 h on dioestrus produced the pattern of response characteristic of dioestrus the next morning, but with much higher levels of LH release, which were unaltered by the addition of E to the medium. OVX at 17.00 h on dioestrus produced an entirely different pattern of response the next day, with high basal and moderate LHRH-induced LH release, and no evidence of changes in sensitivity of the tissue. In this case, E in vitro had a moderate enhancing effect. Various doses of E were injected subcutaneously immediately after OVX at 10.00 h on dioestrus; 1 µg/100 g body weight E had no effect on the subsequent pattern of response (sensitisation followed by desensitisation), but produced significantly greater LH release during the second, third and fourth volleys of LHRH; the pattern of LH release was also unaffected by P alone (2 mg per rat at 17.00 h after OVX), but the output throughout the perifusion was significantly reduced; 10 µg E completely changed the pattern, to one resembling that seen from tissues removed from intact pro-oestrus rats, but basal and LHRH-induced release were still markedly raised. 30 or 100 µg E reduced basal LH release to intact or sham-operated levels, but had no inhibitory effect on the high LHRH-induced release of LH. Both the pattern and magnitude of release were only restored to those exhibited by pituitaries from normal pro-oestrous animals when both 10 µg/100 g E and 2 mg P were given. In conclusion, these results suggest that this preparation is a suitable in vitro model for studying the mechanisms involved in the LH surge. Prolonged sensitisation to repeated LHRH stimulation is characteristic of pro-oestrous pituitaries, whereas dioestrus tissues show desensitisation. The pro-oestrous pattern could be restored in OVX animals by sufficient E replacement for 1 day, the addition of P replacement reducing the output of LH to