To compare the effects of PGF2α (dinoprost tromethamine) and D-cloprostenol in a two-dose protocol for estrus synchronization in hair sheep during breeding season in Yucatán, México, two experiments were conducted. In experiment 1, 61 cyclic hair sheep were divided into two groups: G1 (control n = 30), two doses of 50 μg of dinoprost tromethamine IM with 12 days between applications, and G2 (n = 31), two doses of 50 μg of D-cloprostenol IM at the same time interval. For determination of progesterone levels, 16 ewes from each group were randomly selected. In experiment 2, 70 cyclic hair sheep were assigned at the same treatments (G1 and G2, n = 35) and 48 h after the second application, the ewes in estrus were detected by two vasectomized rams. Sheep with detected estrus were inseminated, and 45 days after, pregnant animals were identified by ultrasonography. An exact Fisher's test was performed for the analysis of ewes in estrus (experiments 1 and 2) and number of pregnant ewes (experiment 2); for the comparison of time between end of treatment-estrus presentation, a survival analysis was used. Duration of estrus in hours was analyzed using a generalized mixed model (GLM) ANOVA whereas plasma progesterone concentrations were analyzed by non-linear regression. There were significant differences (P < 0.05) in the proportion of ewes in estrus upon treatments (G1, 57% vs G2, 87% and G1, 37.1% vs G2, 65.7% in experiments 1 and 2, respectively), and between the end of treatment-onset estrus interval (P < 0.01), survival curves showed the highest number of sheep in estrus between 40 and 48 h (G1, 43.7 + 8.05 h vs G2, 42.9 + 6.7 h, experiment 1). There were no significant differences (P > 0.05) in duration of estrus (G1, 42 + 6.1 h, vs G2, 41.1 + 11.2 h, experiment 1) and pregnancy in the ewes that presented estrus, and were inseminated (G1, 38.4% vs 52.1%, experiment 2). With regard to concentrations of progesterone, significant differences (P < 0.01) were found between treatments, and progesterone levels before the second application of D-cloprostenol were higher. In consideration of the results, it can be concluded that in a two-dose protocol of a luteolytic agent, more ewes presented estrus in response to D-cloprostenol compared to dinoprost tromethamine with similar pregnancy rates.