This meta-analysis aimed to compare the outcomes of the gonadotrophin-releasing hormone (GnRH) antagonist/ letrozole protocol with those of the conventional GnRH antagonist protocol for poor responders undergoing in vitro fertilization/intracytoplasmic sperm injection (IVF/ICSI). We searched for relevant articles in PubMed, EMBASE, Google Scholar, and retrieved 452 records. Eventually, we selected five eligible trials with data for 564 patients characterized as poor ovarian responders. Our meta-analysis revealed that the clinical pregnancy rate (per cycle) with administration of letrozole might be a higher than that in the control groups (risk rate [RR]: 1.57, 95% confidence interval [CI]: 1.00-2.44, p = 0.05). .Moreover,it indicated that the total dose of gonadotrophin was significantly decreased with the administration of letrozole compared to control groups(mean difference [MD]: -529.37, 95% CI: -1207.45 to -111.25, p = 0.001),.However, there was no statistical difference in the number of retrieved oocytes(MD: 0.59, 95% CI: -0.36-1.54, p = 0.22), cycle cancelation rate (RR: 0.81, 95% CI: 0.58-1.12, p = 0.20), or estradiol concentration on the day of HCG administration(MD: -28.19, 95% CI: -77.71-21.33, p = 0.26) in the presence or absence of letrozole combination in the GnRH antagonist protocol. In conclusion, letrozole administration might improve clinical pregnancy rate in conventional GnRH antagonist protocol for poor responders. Moreover, letrozole co-treatment aslo can reduce the economic burden of poor responders during the GnRH antagonist cycle. Nevertheless, large-scale and multi-center randomized controlled trials are needed to further evaluate the efficacy of adjunctive letrozole administration in the GnRH antagonist protocol. (C) 2021 Elsevier Masson SAS. All rights reserved.