The best available evidence on surgery for endometriosis-associated pain has been reviewed in order to define the benefit of various interventions in the most frequently encountered clinical conditions, and discuss the robustness of the reported data in light of the quality of the relevant study design. Methodological drawbacks limit the validity of observational, non-comparative studies on the effect of laparoscopy for stage I to IV disease. The results of three randomized, controlled trials, indicate that the absolute benefit increase of destruction of lesions compared with sham operation in terms of proportion of women reporting pain relief was between 30% and 40% after short follow-up periods. The effect size decreased with time and the reoperation rate, based on long-term follow-up studies, was as high as 50%. In most case series on excisional surgery for rectovaginal endometriosis, substantial short-term pain relief was experienced by about 70-80% of the subjects who continued the study. However, at one-year follow-up approximately 50% of the women needed medical treatments. Major complications were observed in 3-10% of the patients. Medium-term recurrence of lesions was observed in about 20% of the cases, and around 25% of the women underwent repetitive surgery. Routine complementary performance of denervating procedures cannot be recommended based on the quality of the available information, as only a few symptomatic patients complain of exclusively midline, hypogastric pain. Pain recurrence and reoperation rates after conservative surgery for symptomatic endometriosis are high and probably underestimated. Clinicians and patients should be aware that the expected benefit is operator-dependent and, especially in complex conditions, acceptable results can be assured in referral centers.