Objective To compare the surgical techniques and clinical outcomes between transvesical and posterior approaches for Retzius-sparing robot-assisted laparoscopic radical prostatectomy (RARP). Methods The clinical data of 35 patients with prostate tumors receiving Retzius-sparing RARP between Dec. 2016 and Dec. 2018 were retrospectively analyzed. Twenty-two cases underwent Retzius-sparing RARP via posterior approach (pRARP), and the preoperative data (total prostate-specific antigen [tPSA] [16.9±7.5] ng/mL, Gleason score 7 [5-8], prostate volume [42.3±11.2] mL and international index of erectile function-5 [IIEF-5] score 14 [9-20]) were documented, with nine cases at cT1c stage, four at cT2a, three at cT2b and six at cT2c. There were 13 cases in transvesical RARP (TvRARP) group, and the preoperative data (tPSA [18.6±8.7] ng/mL, Gleason score 6 [5-7], prostate volume [35.4±9.6] mL and IIEF-5 score 15 [10-21]) were also recorded, with eight cases at cT1c stage, two at cT2a, two at cT2b and one at cT2c. All patients had urinary continence preoperatively. Both pRARP and TvRARP were performed transperitoneally. The operation time, intraoperative blood loss, postoperative staging, postoperative Gleason score, recovery of urinary continence and postoperative IIEF-5 score were compared between the two groups. Results The operations were successfully completed in all the 35 cases without conversion to open surgery, blood transfusion or other major surgical complications. The operation time were (123.4±31.7) min and (135.3±25.6) min, and the estimated blood loss were (48.2±12.8) mL and (65.2±19.8) mL in pRARP and TvRARP groups, respectively, without significant differences (both P>0.05). Postoperative pathology confirmed six pT2a cases, six pT2b cases, eight pT2c cases and two pT3a cases in pRARP group, with Gleason score being 7 (5-8); and eight pT2a cases, three pT2b cases and two pT2c cases in TvRARP group, with Gleason score being 6 (5-7) (both P>0.05). Positive surgical margins were found in five cases (22.7%) in pRARP group and three cases (23.1%) in TvRARP group (P>0.05). Urethral catheter was removed at postoperative 7-day in both groups. Eighteen patients in pRARP group and 12 patients in TvRARP group achieved urinary continence (without pads) immediately after the removal of urinary catheter, while four patients in pRARP group and one patient in TvRARP group achieved urinary continence 2 weeks postoperatively (P>0.05). IIEF-5 score was 11 (4-13) in pRARP group and 12 (5-14) in TvRARP group 2 months postoperatively (P>0.05). There was no significant difference in IIEF-5 scores before and after operation in the two groups (both P>0.05). During a follow-up of 3-24 months, no biochemical recurrence was noted (tPSA<0.2 ng/mL). Conclusion The clinical outcomes of Retzius-sparing TvRARP for small localized prostate cancer are similar to those of Retzius-sparing pRARP, with satisfactory immediate postoperative urinary continence. Though the preservation of erectile function requires further follow-up, Retzius-sparing TvRARP is technically feasible and might be a valid alternative for localized low-risk prostate cancer.