Background: Drug cost analysis has not been performed to compare different anesthetic techniques used in prolonged procedure surgery under auditory evoked potential (AEP) monitoring. We compared the costs of propofol-based total intravenous anesthesia (TIVA), sevoflurane (SEVO) anesthesia and desflurane (DES) anesthesia in prolonged lumbar spine surgery lasting more than 4 h. Methods: We studied 75 patients (ASA class I-II) scheduled to undergo elective lumbar spine surgery. Patients were randomized into TIVA, SEVO, or DES groups. The AEP index was maintained at 15-25. After completing surgery, emergence time from anesthesia was measured and the anesthetic costs were calculated in New Taiwan dollars (NTD). Results: Demographic factors were not significantly different between groups. The total cost was significantly higher in the SEVO group than in the TIVA and DES groups (NTD 2052±485, 1327±316 and 1439±313, respectively; P＜0.001). The main anesthetic drug was significantly cheaper in the TIVA than in the SEVO and DES groups (NTD 846±216, 1616±456 and 969±304, respectively; P＜0.001). The cost of perioperative fentanyl and cisatracurium were higher in the TIVA group than in the SEVO and DES groups. The TIVA group had faster recovery than the DES and SEVO groups (9.7±4.0, 26.8±11.9 and 23.1±11.6 min, respectively; P＜0.001). Conclusions: Patients undergoing prolonged procedure recover faster after TIVA than after DES or SEVO anesthesia. The TIVA and DES regimens were significantly cheaper than SEVO-based anesthesia.