The objective of this study was to compare the reproductive performance of lactating dairy cows using a treatment (TRT) program for second and subsequent artificial insemination (AI) services aimed at (1) increasing AI upon estrus detection based on increased physical activity (AIAct) and (2) increasing fertility of timed AI (TAI) services for cows not AIAct through presynchronization of the estrous cycle and improved physiological milieu before TAI. Cows in the control (CON) group were managed with a program that combined AIAct and TAI after the Ovsynch protocol. After nonpregnancy diagnosis (NPD) by transrectal ultrasonography at 31±3d after AI, cows received the following treatments: (1) CON (n=634), AIAct any time after a previous AI and resynchronization with the Ovsynch-56 protocol (GnRH-7d-PGF2α-56 h-GnRH-16 h-TAI) 1d after NPD, or (2) TRT (n=616): cows with a corpus luteum (CL) ≥20mm (TRT-CL) received a PGF2α injection 1d after NPD, whereas cows with no CL or a CL <20mm (TRT-NoCL) received a GnRH injection 3d after NPD. Cows in TRT-CL and TRT-NoCL not AIAct were enrolled in a 5-d Ovsynch + progesterone protocol (GnRH + controlled internal drug release-5d-PGF2α + controlled internal drug release removal-24 h-PGF2α -32 h-GnRH-16 h-TAI) 9 and 7d after the PGF2α or GnRH injection, respectively, to receive TAI. The hazard of pregnancy up to 270 DIM was similar for cows in the CON and TRT group (hazard ratio=1.07, 95% CI=0.95 to 1.21), but it was affected by parity (primiparous greater than multiparous cows). Median days to pregnancy for the CON and TRT group were 111 and 110d, respectively. When evaluated after 104 DIM (first time point at which cows were affected by the treatments), the hazard of pregnancy was similar for the CON and TRT group (hazard ratio=1.15, 95% CI=0.95 to 1.39). Based on this analysis, median days to pregnancy for the CON and TRT group were 161 and 178d, respectively. Thus, in spite of increasing the proportion of cows AIAct (29 and 10% for TRT and CON), median days to insemination after NPD were greater for cows in the TRT (17d) than the CON (10d) group, which coupled with similar fertility to AIAct, and TAI failed to improve overall reproductive performance. A low proportion of cows with a CL at NPD (65.2%) and a poor response to PGF2α may explain the poor estrus detection efficiency in the TRT group. We concluded that, when compared with a typical estrus detection and TAI program for cows failing to conceive to previous AI services, a program aimed at increasing the proportion of cows AIAct after NPD and fertility of TAI services increased the proportion of cows AIAct but failed to reduce days to pregnancy during lactation because of greater days to AI after NPD.