BACKGROUND: Whether monitored anesthetic care (MAC) under total intravenous anesthesia can substitute for spinal anesthesia in knee arthroscopic surgery in regard to frequency of postoperative complications, degree of postoperative pain and degree of the satisfaction of patients and operator was investigated. METHODS: Sixty healthy patients were allocated randomly into a spinal group (n = 30) who received spinal anesthesia and an MAC group (n = 30) who received TIVA for anesthesia for arthroscopic surgery. All patients were NPO for 8 hours before surgery were premedicated and monitored with an EKG, noninvasive blood pressure and pulse oximeter. Heavy 0.5% bupivacaine, 10 - 12 mg, was used for spinal anesthesia and fentanyl 2ng/kg, propofol 1 mg/kg, ketamine 0.3 mg/ kg, and ketorolac 30 mg were given intravenously for induction and propofol was maintained at 3 - 4 mg/kg/h for TIVA. Local anesthetics infiltration was done at the arthroscopic portal site and fentanyl 25ng and propofol 20 mg were added intermittently. Postoperative complication (nausea, vomitting, back pain, dizziness, pain or voiding difficulty) and satisfaction of the patients and surgeon were investigated by VAS and 5 grade methods, respectively. RESULTS: The frequency and degree of back pain, pain at the operative site and voiding difficulty occured less and the VAS was decreased in the MAC group while in the recovery room, 6 hours and the day after the operation. The grade of satisfaction of the patients and that of the surgeon were also high in the MAC group. CONCLUSIONS: These results show that, if careful airway management is provided, MAC under TIVA is a more useful anesthetic method than spinal anesthesia in an arthroscopy.