Approximately 60% of lactating Holstein dairy cows fail to conceive to the first artificial insemination (AI) and are required reinsemination. Resynchronization facilitates timely reinsemination; however, pregnancy rates to the second AI (PR/AI) from resynchronization protocols have been inconsistent. The development of an effective resynchronization protocol is needed to improve PR/AI to second AI. The objective of this study was to determine the effect of the initial GnRH on PR/AI using a 5-d CIDR-Cosynch resynchronization protocol for second AI in lactating Holstein dairy cows. Approximately 37 d after first AI and upon non-pregnancy diagnosis, cows eligible for second AI (n = 429) were subjected into a modified 5-d CIDR-Cosynch protocol. On d 0, all cows received CIDR inserts and were assigned randomly to oe one of 2 treatments, initial GnRH (100μg; n = 226) or no GnRH (n = 203). On d 5, CIDR inserts were removed and all cows received one injection of PGF_2α (500 μg). From d 5 to 7, cows were monitored daily for estrus. Animals detected in estrus on d 6 or 7 received AI and were not given GnRH. On d 8, all cows not identified in estrus on d 6 or 7 were timed AI (TAI) and receieved GnRH. Blood samples were collected from cows (n = 184) on d 0 and analyzed for progesterone (P₄) concentration. Logistic regression was used to examine the main effects of treatment, day of AI, sire, parity and all 2-way interactions with treatment on PR/AI. There were no effects of treatment (no GnRH, 27.1% vs. initial GnRH, 21.2%) on PR/AI (P > 0.10). There were also no effects of method of AI (based on estrus vs. TAI), sire (n = 10), or any 2-way interactions with treatment on overall PR/AI. There was a significant difference (P < 0.05) between primiparous (n = 119) and multiparous cows (n = 310) in PR/AI (31% vs. 21%). Progesterone concentrations (n = 184) were categorized into high (≥1 ng/mL) and low (<1 ng/mL) groups. Overall, PR/AI tended (P = 0.09) to be greater for high P₄ (n = 136) compared with low P₄ (n = 48) (26% vs. 16%). These results provide evidence that the initial GnRH injection in a 5-d CIDR-Cosynch resynchronization program may not be necessary to achieve a similar PR/AI following second AI.