The aim in this study was to compare two estrus synchronization protocols in buffaloes. Animals were divided into two,groups: Group A (n = 111) received 100 mug GnRH on Day 0, 375 mug PGF(2alpha) on Day 7 and 100 mug GnRH on Day 9 (Ovsynch); Group B (n = 1] 7) received an intravaginal drug release device (PRID(R)) containing 1.55 g progesterone and a capsule with 10 mg estradiol benzoate for 10 days and were treated with a luteolytic dose of PGF(2alpha) and 1000 IU PMSG at the time of PRID(R) withdrawal. Animals were inseminated twice 18 and 42 h after the second injection of GnRH (Group A) and 60 and 84 h after PGF(2alpha) and PMSG injections (Group B). Progesterone (P-4) concentrations in milk samples collected 12 and 2 days before treatments were used to determine cyclic and non-cyclic buffaloes, and milk P4 concentrations 10 days after Artificial insemination (AI) were used as an index of a functional corpus luteum. Cows were palpated per rectum at 40 and 90 days after AI to determine pregnancies. All previously non-cyclic animals in Group B had elevated P-4 (>120 pg/ml milk whey) on Day 10 after AI. Accordingly, a greater (P < 0.01) relative percentage of animals with elevated P4 10 days after AI were observed in Group B (93.2%) than in Group A (81.1%). However, there was no difference in overall pregnancy rates between the two estrus synchronization protocols (Group A, 36.0%; Group B 28.2%). When only animals with elevated P4 on Day 10 after AI were considered, pregnancy rate was higher (P < 0.05) for animals in Group A (44.4%) than Group B (30.3%). The findings indicated that treatment with PRID(R) can induce ovulation in non-cyclic buffalo cows. However, synchronization of estrus with Ovsynch resulted in a higher pregnancy rate compared with synchronization with PRID(R), particularly in cyclic buffalo. (C) 2002 Elsevier Science Inc. All rights reserved.