We attempted to compare the analgesic effects of tramadol infusion intravenously and epidurally administered through a patient-controlled analgesia (PCA) method for postoperative analgesia following gynaecological cancer surgery. Forty patients undergoing elective cancer surgery, included in the American Society of Anesthesiologists (ASA) class II and III, were randomly placed into two groups. The patients in the intravenous (IVA) group were administered a 20 mg bolus of tramadol intravenously and the patients in the epideral analgesia (EA) group epidurally five minutes before induction. The PCA equipment was programmed to deliver 20 mg of tramadol as a bolus dose, with a lock-out time of 15 minutes, at a 10 mg/hour infusion rate in both groups. A visual analogue scale (VAS) and patient satisfaction as well as haemodynamic and respiratory parameters were determined at given times postoperatively. Total tramadol consumption at 24 hours and side-effects were recorded. There was no difference between groups based on haemodynamic and respiratory parameters whereas there was a significant difference based on tramadol consumption, VAS and side-effects of tramadol and patient satisfaction between groups. VAS values of patients, 6.85 +/- 1.34 and 3.00 +/- 1.58, respectively, for the IVA group (group 1) and the EA group (group 2) were found to be significantly different. Postoperative patient satisfaction was higher was in group 2 than in group 1 (3.45 and 2.7, respectively). In conclusion, epidural administration of tramadol through the PCA method following gynecologic cancer surgery was found to be a more effective analgesia in lower doses when compared to the intravenous administration.