The effect of breed and age on the concentration of anti-mullerian hormone (AMH) and the relationship between AMH and fertility in replacement ewes were evaluated. In Experiment 1, a single blood sample was used to compare concentration of AMH in Dorset/Texel (DT; n= 238; age 8.7 ± 0.1 months), Suffolk (n= 44) and Katahdin (n= 77; age 6.9 ± 0.04 months) replacement females and to determine changes in systemic AMH with age in DT and Katahdin females. In Experiment 2, Katahdin and DT females were placed into LOW, MEDIUM and HIGH groups based on their systemic AMH determined from a blood sample collected 2 months prior to breeding. Females were treated with CIDR inserts (0.3g progesterone) for 5 days and were exposed to rams at insert removal for 30-35 days. Ewes were observed for estrus after 4 days of ram exposure, and pregnancy diagnosis was conducted via transrectal ultrasonography at the time of ram removal and again 20-25 days. In Experiment 1, Katahdin females had a higher AMH than DT and Suffolk females (P < 0.001: 566 ± 37 vs. 337 ± 14 vs. 237 ± 22 pg/ml, respectively). AMH decreased linearly with age in DT females (P = 0.03). In Experiment 2, Females with high AMH conceived and lambed to the first service in Katahdin but not the DT breed (Breed x AMH; P < 0. 05). Replacement females that conceived (Breed X Conception; P < 0.001) and lambed to the first service (Breed X lambing to 1 st service; P < 0.001) had a higher AMH in Katahdin but not DT breed. In conclusion, AMH varies among different breeds of sheep and a single measure may be useful to select females with high reproductive performance in some breeds.