Objective To establish a virtual reality technology-based segmented training system and apply it to the surgical training of robot-assisted laparoscopic pyeloplasty (RALP). Methods Virtual reality technology was used to conduct basic operation training and intensive training of suture-related skills for 3 urologists through a robotic simulation training system, and the changes in key indicators such as anastomosis time of the trained urologists were compared before and after the training. The RALP was divided into 4 segments: separating, cutting, anastomosis, and suturing. The trained surgeons participated in real surgical operations through the robotic system, first completing the learning curve in segments and then completing the whole RALP independently. Results After virtual reality training, the anastomotic time (［141.0±9.0］ s vs ［312.6±34.5］ s) and motion distance (［140.9±16.2］ cm vs ［323.1±50.6］ cm) were significantly shortened (both P＜0.01), and the times of both instrument collisions (1.6±0.6 vs 7.1±1.0) and instrument out of sight (0.8±0.5 vs 3.2±0.9) were significantly reduced (both P＜0.05). However, the times of missed target did not change significantly before and after training (P＞0.05). After 2 rounds of segmented training per person, trainees were able to conduct RALP independently. Their operation time was longer than that of senior surgeons (［149.0±36.0］ min vs ［117.0±10.0］ min), but the difference was not statistically significant (P＞0.05). In addition, there were no significant differences in perioperative indicators such as intraoperative blood loss, incidence of urine leakage, or postoperative hospital stay (all P＞0.05). Conclusion The virtual reality technology-based segmented training system can be successfully applied to the surgical training of RALP, and it can improve the operational skills of trainees on the premise of ensuring the safety and quality of the operation.