Five groups of 30 captive feral mares each were implanted with silastic rods containing estradiol (E) and/or progesterone (P): E only with 8 g, P only with 24 g, P HE with 8 g P 8 g E, HP E with 12 g P 4 g E, HP LE with 12 g P 2 g E. Arbitrary group designations were differentiated by relative high (H) and low (L) amounts of steroid. Thirty mares received silastic rods containing no hormone (CI). Five mares from each group were bled every 2 wk for 4 mo and monthly for another 5 mo. All mares were tested for estrus by allowing them to stand in an alley between two pens of stallions and visually monitoring her response to the stallion. Serum P levels increased from 0.3 ± 0.1 to 1.8 ± 0.1 ng/ml in the P only group during the first 3 wk after implanting. Levels remained stable for the next 2 wk and then began a gradual decline. Serum P levels in the other groups were lower. Serum E levels were slightly increased in the groups receiving 8 g of E (E only and P HE groups). Significantly fewer animals in the E only and P HE groups exhibited estrus as compared with control animals (10 of 23 and 13 of 26 versus 22 of 25, respectively, P ≤ 0.003). However, animals receiving 24 g of P (P only) showed similar occurrences of estrus as controls. All animals detected in estrus ovulated as evidenced by elevations in serum P levels above 5 ng/ml collected 10 days after the mare was detected in heat. In spite of the significant effect of hormone implants on the occurrence of estrus, over 80% of mares bred and conceived when placed with a stallion. Our data suggest that these levels of implanted hormones can raise serum levels of P and E for at least 21 wk. These hormones, when implanted during early anestrus, can alter the occurrence of psychic estrus in captive feral mares without suppressing ovulation and conception.