Although protocols for superovulation have been described in horses, this technique has been discouraged due to the low embryo recovery rates in superovulated mares. The reason for these poor results is poorly understood, but the formation of a blood clot in the ovulation fossa following ovulations has been hypothesized. Therefore, this study aimed to assess the safety and effect of periovulatory anticoagulant therapy on embryo recovery of superovulated mares. In experiment 1, five mares were assigned to receive five anticoagulant treatments in a crossover design: intravenous injections of 150 (H1), 300 (H2), 400 (H3), 450 (H4), 600 (H5) IU/kg of unfractionated heparin (UFH, heparin sodium); and had blood samples sequentially collected for up to 48 h post-treatment to test Prothrombin (PT) and activated partial thromboplastin time (aPTT). In experiment 2, four mares were treated in a crossover design with intravenous injection of 450 IU/kg of UFH and 1 mg/kg of low molecular weight heparin (LMWH, enoxaparin) and had blood collected as previously for analysis of plasma anti-Xa activity. In experiment 3, eleven mares had four cycles randomly assigned to four groups. In the control group, mares did not receive any treatment. In contrast, in groups G1, G2, and G3, mares were superovulated with equine pituitary extract and treated 34 h after the induction of ovulation with a placebo (NaCl 0.9 %, G1), 450 IU/kg of UFH (G2), or 1 mg/kg of LMWH. Mares in all groups had ovulation induced with hCG plus histrelin acetate and were bred with fresh semen from one stallion. Embryo flushing was performed nine days post-ovulation. In experiment 1, only mares in groups H4 and H5 had increased aPTT and PT for up to 12 h, and in all groups, aPTT and PT values returned to baselines at 24 h post-treatment. In experiment 2, plasma anti-Xa activity was increased by both therapies for up to 12 h after treatment and was at baseline levels 24 h post-treatment. In experiment 3, periovulatory therapy with anticoagulants increased embryo recovery rates per cycle (G2, 250 %; G3, 260 %) compared to control-assigned cycles (60 %; P < 0.05), whereas G1-assigned cycles (160 %) had intermediate embryo recovery. In conclusion, periovulatory anticoagulant therapies may be an alternative to improve embryo recovery in superovulated mares.