OBJECTIVE The objective of this study was to assess the interobserver variability of pelvic examination under anesthesia (EUA) in cervical cancer. METHODS Subsequent patients undergoing a staging procedure under anesthesia for primary cervical cancer were enrolled in the study. All clinicians assessed "blinded" tumor size, and the involvement of vagina, parametria, sacrouterine ligaments, pelvic sidewalls, bladder, and/or rectum. Items were scored varying from 1 ("certainly no involvement"), 2 ("not sure about involvement"), to 3 ("involvement"). Each individual decided on the International Federation of Gynecology and Obstetrics (FIGO) stage; also, the urge for imaging and treatment proposal were accounted for. Final FIGO staging was obtained by consensus of the team. Investigators were classified as experienced after more than 50 EUAs. All others were classified less experienced. The free-marginal κ values between experienced and less experienced investigators were calculated for all previously mentioned items. RESULTS Between February 2009 and December 2010, a total of 86 patients were enrolled. Among experienced investigators, a moderate interobserver agreement was found with regard to FIGO stage (free-marginal κ value of 0.49) and an excellent interobserver agreement on their proposed therapy (free-marginal κ value of 0.84). A lower level of agreement was found when comparing experienced with less experienced investigators: only a slight level of agreement on FIGO stage and a substantial agreement on their therapy proposal (free-marginal κ values of 0.03 and 0.66). CONCLUSIONS We describe only a moderate interobserver agreement on clinical staging of patients with cervical cancer. The interobserver agreement increases in the group of experienced doctors, indicating that EUA can be learned.