To determine the outcome of laparoscopy compared with laparotomy in conservative surgical treatment for severe endometriosis. Comparison of nonrandomized historical surgical series. Two teaching hospitals and referral centers specializing in reparative and reconstructive surgery. A total of 216 patients operated for severe endometriosis during a 5-year period. Conservative surgical treatment at laparoscopy (n=67) or laparotomy (n=149) with median follow-up of 24 months. Cumulative probability of pregnancy in previously infertile patients (22 in the laparoscopy group and 70 in the laparotomy group) and cumulative probability of pain recurrence in subjects with moderate or severe symptoms before surgery (47 in the laparoscopy group and 108 in the laparotomy group). The 24-month cumulative probability of pregnancy according to the Kaplan-Meier method was 44.9% after laparoscopy and 62.7% after laparotomy. The 24-month cumulative probability of symptoms recurrence evaluated with a 0 to 3 point verbal rating scale was, respectively, 16.4% versus 20.3% for dysmenorrhea, 33.3% versus 15.4% for deep dyspareunia, and 25.0% versus 15.9% for nonmenstrual pain. The corresponding figures obtained with a 10-point linear analogue scale were 20.3% versus 24.7%, 28.6% versus 10.4%, and 17.5% versus 20.1%. No difference is statistically significant. Laparoscopy and laparotomy seem equally effective in the treatment of infertility and chronic pelvic pain associated with severe endometriosis. However, a trend was observed toward a higher pregnancy rate and lower dyspareunia recurrence rate after surgery for severe endometriosis performed at laparotomy compared with laparoscopy.