Drug resistance and severe pelvic pain often warrant surgical intervention for treating deep endometriosis (DE); however, damage to the autonomic nervous system can occur because of anatomical considerations. We aimed to investigate the advantages of robotic technology in enabling precise dissection, even in DE. We retrospectively compared the surgical outcomes of robot-assisted (RA) and conventional laparoscopic (CL) nerve-sparing modified radical hysterectomies (NSmRHs) for DE. Between the two groups (RA-NSmRH group, n = 50; CL-NSmRH group, n = 18), no differences were identified based on patient demographics, such as age, body mass index, previous surgery, revised American Society of Reproductive Medicine classification, Enzian classification, uterine weight, number of removed DE lesions, and concomitant procedures. All patients in both groups achieved complete removal of the DE lesions with complete bilateral pelvic autonomic nerve preservation. The mean operative time (OT) was significantly longer (130 ± 46 vs. 98 ± 22min, p < 0.01), and estimated blood loss (EBL) was lower (35 ± 44 vs. 131 ± 49ml, p < 0.01) in the RA-NSmRH group than in the CL-NSmRH group. The hospitalization days (4.3 ± 1.3 vs. 4.1 ± 0.2days, p = 0.45) and perioperative complications with Clavien-Dindo classification ≥ grade III (0% vs. 0%) were not significant in both the groups. None of the patients required self-catheterization after surgery. Compared with CL-NSmRH, RA-NSmRH was associated with longer OT and lower EBL, whereas the number of hospitalization days and complications were similar in both groups. Our results imply that nerve-sparing surgery can be safely and reproducibly performed using conventional or robotic laparoscopic modalities to treat DE.