Thirty-seven seasonally anoestrous mares were divided into treatment and control groups and given 10 micrograms of native GnRH (GnRH) per hour using a peristaltic pump, or 10 micrograms GnRH agonist (GnRHa) twice daily, beginning on either 13 January, 13 February or 14 March. Treatment with GnRH was equally effective in inducing ovulation in January (4/5), February (4/5) and March (3/4). GnRHa treatment was more effective in inducing ovulation in February (4/5) and March (4/4) than in January (2/8). Peak luteinizing hormone (LH) concentrations in mares induced to ovulate with GnRH (7.4 +/- 1.5 ng/ml) were significantly higher than LH concentrations in mares induced to ovulate with GnRHa (1.8 +/- 0.2 ng/ml). Urinary oestrogen conjugate concentrations increased parallel to increases in follicular diameter during treatment. Ovulations induced by GnRH or GnRHa were followed by a normal luteal phase. All mares induced to ovulate in January and February returned to anoestrous following withdrawal of GnRH support. The results suggest that the efficacy of GnRHa in the induction of ovulation in anoestrous mares is influenced by season, whereas the efficacy of pulsatile GnRH administration is not affected by season.