To investigate the hypothesis that the onset of the breeding season in the mare may be due to a daylength-induced seasonal increase in LHRH pulse frequency, 5 mares were immunized against LHRH. Beginning 1 December, 5 immunized and 5 untreated control mares were exposed to an abrupt, artificial increase in daylength (16L:8D) to advance the onset of the breeding season. In control mares ovulation occurred 49.6 +/- 3.5 (s.e.m.) days later (18 January), whereas in 3/5 immunized mares ovulation had not occurred by 1 April. In the remaining 2 mares, although ovulation occurred once (Mare 79) or twice (Mare 72) during February, a booster immunization restored acyclicity for the duration of the study (No. 72) or to 30 March (No. 79). The absence or occurrence of ovulation in LHRH-immunized mares appeared to be related to antibody titre, such that the highest antibody titres were observed in those mares that remained anovulatory throughout the experimental period, while low titres were seen in the 2 mares that ovulated. In 3 acyclic immunized mares, LH pulses were not observed in blood samples collected frequently at 2-week intervals from November to January. In contrast, LH pulse frequency in control mares, and in the 2 immunized mares that ovulated, increased from 0/12 h (November) to 2-5/12 h (January). These results confirm our previous observation that, in the mare, the onset of the breeding season is associated with an increase in LH pulse frequency. Furthermore, the results suggest that the increase in LH pulse frequency reflects an increase in pulsatile LHRH release from the hypothalamus.