Data on the estrous cycles and sexual receptivity scores of 7 maiden Australian Stock Horse mares were used to study seasonal variation from December until the following November. Mares were grazed in paddocks containing both native and improved tropical pasture species. The study was conducted in South-East Queensland (27°27′ South latitude). There were no differences between either the sexual receptivity or the lengths of natural (Mean=7.5 days, SEM=0.4) and PGF-induced estruses (Mean=7.6 days, SEM=0.4). Plasma estrogens were not related to differences in sexual receptivity scores between mares. The lowest incidence of estrus occurred at the time of the winter solstice (June 22) in the Southern Hemisphere. Winter estruses (Mean=9.3 days, SEM=0.8) were longer than those of summer (Mean=6.6 days, SEM=0.5) and autumn (Mean=6.6 days, SEM=0.9). Approximately 30% of estruses were anovulatory. Most of these occurred in autumn and winter (P<0.05). There was considerable variation in the occurrence of anovulatory estrus between mares. There were no differences between the lengths of ovulatory and anovulatory estruses. Ovulatory estruses were significantly longer in spring than in summer (P<0.05). There were no seasonal differences between the lengths of anovulatory estruses. The length of diestrus (Mean=16.6 days, SEM=0.3) was not affected by mare or season. Only 3 of the 7 mares did not cycle during winter (anestrus), as confirmed by progesterone concentrations of less than 1 ng/ml. Finally, there was no evidence of mares having 2 breeding seasons per year in this study.