Purpose: We evaluated the efficacy and safety of a mentor-initiated program for laparoscopic radical prostatectomy by analyzing its effect on the learning curve. Patients and Methods: The mentor performed 16 procedures (group I) and the trainee, assisted by the mentor, 12 (group II). The next 16 procedures were performed by the trainee without the mentor (group III). The patient groups were comparable in terms of age, serum prostate specific antigen concentration, Gleason score, and clinical stage. The operating time, blood loss, complications, and outcomes were evaluated. Statistical analysis was performed using ANOVA with the multiple-comparisons test with Bonferroni correction and the Kruskal-Wallis test, when appropriate. Results: There was a statistical difference in the mean operating time in groups I and II (271 and 381 minutes, respectively; P < 0.001) and in groups I and III (271 and 386 minutes, respectively; P < 0.001), but the difference between groups II and III was not significant (P > 0.05). The mean estimated blood loss was similar in all groups (362, 395, and 434 mL, respectively; P = 0.86). The mean postoperative day 1 decrease in hemoglobin was similar in the three groups (0.65, 0.66, and 0.66 mg/dL, respectively; P = 1.00). No patient required open conversion. Postoperative complication rates were the same in groups I and III (6.25%). The mean catheterization time was longer in group III (6, 7, and 12 days; P < 0.001). The mean hospital stays (9, 8, and 8 days; P = 0.28) were similar. Stage pT3-pT4a disease was found in 75%, 41.6%, and 75% of the specimens in groups I to III, respectively. There was no statistical difference in positive-margin rates in the three groups (43.8%, 33.3%, and 37.5%; P = 0.85). Conclusions: A mentored program allows safe introduction of laparoscopic radical prostatectomy into surgical practice. Nevertheless, during the learning curve, longer operating and catheterization times have to be expected.