BACKGROUND: Because of wide individual variations in response to sedative and the level of sedation desired by different patients, inadequate sedation is frequent during surgery. Patient-controlled sedation is a logical extension of patient-controlled analgesia to find and maintain their own steady-state of sedation by self-administration of sedatives during surgery. The purpose of this study was to evaluate the feasibility of patient-controlled sedation compared with anesthesiologist-controlled sedation during surgical spinal anesthesia. METHODS: Unpremedicated forty adult patients who received spinal anesthesia for lower extremity surgery were randomly allocated into two groups (n=20 for each group). After selection of target state of sedation according to sedation scale, patient-controlled sedation (PCS) group self-administered 0.5 mg (1 ml) intravenous midazolam in increments using a Walkmed PCA infusor and anesthesiologist- controlled sedation (ACS) group administered by the anesthesiologist as the same manner to achieve previously selected sedation state. Sedation score, vital signs, SpO2 were checked 5, 10, 20, 30, 40min after start of drug injection. RESULTS: The sedation scores patient desired were 4.4 +/- 0.8 in PCS group and 4.3 +/- 0.7 in ACS group. These scores were achieved 20min after start of injection in PCS group and 40 min in ACS grou p (p<0.05). Degree of satisfaction was higher in PCS group compared with ACS group (1.5 +/- 0.6 vs 2.1 +/- 0.8, p<0.05). No complications were detected in two groups. CONCLUSIONS: PCS using midazolam was better than ACS in terms of early achievement of sedation state patient desired and degree of satisfaction.