To investigate seasonal effects on the efficacy of estrus synchronization in mares, we administered a progesterone-releasing device (PRID) intravaginally to eight Haflinger mares for 11 days. In January 3 of 8 mares responded to the treatment with estrus and ovulation, in March 7 with estrus and 6 of 7 mares with ovulation, in June 6 of 7 and in October 7 of 8 mares with estrus and ovulation. Follicle distribution patterns at PRID insertion were different between January/October, March/June and June/October (P<0.05). Number of follicles decreased during PRID treatment in January, March and June (difference of number of follicles at Day 12 minus number of follicles at Day 1: -4.2+/-2.7, -0.9+/-0.9 and -4.9+/-1.5 follicles), while it increased in October (3.9+/-1.2 follicles; P<0.05). Mean progesterone concentrations were lowest in January (0.3+/-0.1 ng mL(-1)) when compared with March (3.5+/-1.8 ng mL(-1); P=0.063), June (4.4+/-1.4 ng mL(-1); P<0.05) and October (2.2+/-0.9 ng mL(-1); P<0.05). At Day 2 of PRID treatment, mean progesterone concentrations significantly increased in all mares. Except from January, mean LH concentrations decreased within one day after PRID insertion and remained at low levels during treatments in January and March. Total secretion of LH during PRID-treatment was significantly lower in January and March when compared with June and October. In the 5 of 7 mares that ovulated during PRID treatment a distinct increase of plasma LH concentrations after ovulation was detected. Administration of the progesterone releasing intravaginal device PRID combined with the PGF2alpha analogue cloprostenol was able to induce estrus and ovulation in mares at different times of the year. However, efficacy of the treatment was not satisfactory concerning effectiveness in relation to season and synchrony of intervals from removal of PRID to ovulation in mares.