ABSTRACT The objective of this study was to compare the reproductive efficiency of dairy buffaloes undergoing fixed-time artificial insemination (FTAI) protocols based on progesterone/estrogen (P4/E2) and eCG during unfavorable breeding season using cooled (CS) and frozen semen (FS). A total of 446 buffaloes (> 40 days postpartum) were randomly distributed into four blocks (years): B1-2014 (n = 143), B2-2015 (n = 34), B3-2016 (n = 90), and B4-2017 (n = 179). Each block was subdivided into two (AI with CS and FS using the same ejaculate of each bull). Thus, the block subdivision was as follows: B1 (CS = 71 and FS = 72); B2 (CS = 18 and FS = 16); B3 (CS = 47 and FS = 43); and B4 (CS = 90 and FS = 89). The ejaculates of eight Murrah bulls collected using an artificial vagina were divided into two aliquots: one aliquot was diluted in Botu-Bov® commercial extender and cooled (BB-CS), and the other was diluted in the same extender and frozen (BB-FS). BB-CS aliquots were cooled at 5 °C/24 h using a refrigerator. BB-FS group aliquots were also cooled, and after equilibrating at 5 °C for 4 h, were placed in a 21-L Styrofoam box, 5 cm above the surface of liquid nitrogen. In the afternoon (A) on D0 (2:00 p.m.) the animals received EB 2.0 mg IM (Estrogin®) and an ear implant (CRESTAR® 3.0 mg P4). At D9 (A), the implant was removed, and the animals received eCG 400 IU IM (Folligon® 5000) + Cloprostenol PGF2α 0.530 mg IM (Sincrocio®). At D10 (A), the animals received EB 1.0 mg IM (Estrogin®), and at D12 (8:00 a.m.), AI was performed. At D42, pregnancy was diagnosed via ultrasonography. Total CRs were 48.2% CS and 34.6% FS for years 2014 to 2017, with a significant difference of 13.7% (P<0.05). In conclusion, cooled semen resulted in higher CR than frozen semen in dairy buffaloes under the P4/E2 and eCG FTAI during the unfavorable reproductive season.