A consensus regarding the therapeutic role of lymphadenectomy in endometrial cancer has not been reached because of conflicting negative results of randomized controlled trials and positive results of a cohort study. Since the effects of new treatments tend to be overestimated in observational studies, positive results of an observational study should be validated by a future trial. However, special difficulties are presented in randomized controlled trials in surgery. External validity is important for guaranteeing the reliability of a result of the trial. Physicians' recruitment of eligible patients into a trial depends on the confidence of those physicians for a surgical procedure, workplace environment and feelings of personal responsibility relevant to patients' risk of recurrence. When two surgical procedures are compared in a randomized controlled trial, technical quality control may be reduced in the complicated surgery group due to experienced surgeons' non-participation. It is highly possible that the recruitment issue is a threat to external validity. Therefore, a randomized controlled trial may not be the best format for demonstrating the full benefits of complicated surgery. Multiple studies have demonstrated that the results of well-designed observational studies can be reliable and are comparable with those of randomized controlled trials. Journal editors and funding sources are requested to become more generous with observational studies, especially prospective cohort studies.