〔Abstract〕 Objective To focused on patients with unexplained repeated implantation failure (URIF) and compare the clinical efficacy of different administration routes of granulocyte colony- stimulating factor (G-CSF) in patients undergoing frozen-thawed embryo transfer cycles.Methods A retrospective case-control study was conducted. A total of 187 patients with unexplained recurrent implantation failure (URIF) who received medical treatment were enrolled in this study. According to whether G-CSF was used on the day of endometrial transformation and its administration route in frozen-thawed embryo transfer cycles, the patients were divided into three groups: control group (without G-CSF, n = 71), subcutaneous injection group (subcutaneous G- CSF, n = 72), and intrauterine infusion group (intrauterine G-CSF infusion, n = 44). The pregnancy outcomes among the three groups were compared, and Logistic regression analysis was performed. Results There were significant differences in biochemical pregnancy rates among the three groups (P < 0.05), indicating statistical significance in the overall comparison. In terms of numerical trend, the biochemical pregnancy rate was in the order: subcutaneous injection group > intrauterine infusion group > control group. Further pairwise comparison showed that only the difference between the subcutaneous injection group and the control group was statistically significant (P < 0.017), while no significant differences were found between the other groups (P > 0.017). The overall comparison of early miscarriage rates among the three groups also showed a significant difference (P < 0.05), indicating a statistically significant difference in the distribution among the groups. The numerical trend was: intrauterine perfusion group > subcutaneous injection group > control group. Pairwise comparison results showed that only the difference between the intrauterine perfusion group and the control group was statistically significant (P < 0.017), while there was no statistically significant difference between any other two groups (P > 0.017).Multivariate Logistic regression analysis showed that endometrial thickness on the day of transformation, number of transferred embryos, and subcutaneous G-CSF injection were independent factors associated with an increased biochemical pregnancy rate, where as age was an independent factor for a decreased biochemical pregnancy rate. For clinical pregnancy outcome, endometrial thickness on the day of transformation, number of transferred embryos, and subcutaneous G-CSF injection were also independent factors for an increased clinical pregnancy rate, while BMI was an independent factor for a decreased clinical pregnancy rate. Conclusion Subcutaneous injection of G-CSF only improved the biochemical pregnancy rate in URIF patients. Multivariate Logistic regression analysis showed that subcutaneous G-CSF injection was a favorable factor for clinical pregnancy rate in URIF patients; however, no significant improvement in clinical pregnancy rate was observed among the three groups. Both the comparison of pregnancy outcomes and the results of multivariate logistic regression analysis indicates that intrauterine perfusion of G-CSF has no significant therapeutic effect in the treatment of patients with URIF.