In this study, there was evaluation of pregnancy per AI (P/AI) as a result of the first postpartum AI following four ovulation-synchronization treatments. Treatment regimens were Ovsynch-56 (OVS, n = 875; GnRH1-7d-PGF2α-56h-GnRH2-16h-FTAI), CIDR-OVS (n = 1001; OVS plus CIDR inserts between GnRH1 and PGF2α), Double-Ovsynch (DOVS, n = 663; imposing Pre-OVS followed by breeding-OVS 7 days later), and Modified-DOVS (M-DOVS, n = 1397; imposing Pre-OVS followed either by breeding-OVS or CIDR-OVS based upon the ovarian structure at GnRH1). Cows with a corpus luteum were assigned to a breeding-OVS treatment regimen and those that did not have a corpus luteum were assigned to the CIDR-OVS treatment regimen. Compared with OVS, the P/AI of the cows in the M-DOVS (OR = 1.5, P = 0.001) and CIDR-OVS (OR = 1.4, P = 0.017) was greater at day 30. At day 70, only in the M-DOVS group was there a greater P/AI compared with the OVS group (OR = 1.7, P < 0.001). Pregnancy loss between days 30 and 70 was greater in cows of the CIDR-OVS (OR = 1.9, P = 0.014) compared with those of the OVS group. In cows of the M-DOVS, the dominant ovarian structures (follicle, corpus luteum or cyst) at different time-points of the pre-synchronization period and occurrence of estrus at the end of this period were not associated with P/AI at day 30 post-AI. In conclusion, imposing CIDR-OVS in cows that did not respond to pre-synchronization treatments, resulted in an enhanced pregnancy percentage with the use of the DOVS.