Study Objective: To evaluate the comparative efficacy and side effect profile of ketorolac 60 mg, dezocine 6 mg, and fentanyl 100 μg when used as analgesic supplements to a propofol infusion during monitored anesthesia care (MAC). Design: Randomized, double-blind, placebo-controlled study. Setting: Ambulatory surgery facility at a university medical center. Patients: 80 outpatients undergoing breast biopsy or inguinal herniorraphy procedures under MAC. Interventions: All patients received midazolam 2 mg intravenously (IV) followed by 1 ml of the study medication containing either dezocine 3 mg IV, ketorolac 30 mg IV, fentanyl 50 μg IV, or normal saline. A propofol infusion was initiated at 75 μg/kg/min and then varied to maintain a stable level of sedation ( i.e., Observer Assessment of Alertness/Sedation scale score of 3). An additional 1 ml of the same study medication was administered IV 2 to 3 minutes prior to infiltration of the local anesthetic solution. During the operation, supplemental (rescue) medication consisted of fentanyl 25 μgIV, bolus injections in all four treatment groups. Measurements and Main Results: Propofol infusion and supplemental fentanyl dosage requirements, oxygen saturation values, respiratory rates, recovery times, and postoperative side effects were recorded. Visual analog scales were used to assess sedation, anxiety, pain, and nausea preoperatively (baseline), upon entry into the postanesthesia care unit, and at 30-minute intervals until discharge. The fentanyl and dezocine groups required lower average infusion rates of propofol to maintain a stable level of sedation than the control (saline) group. The saline and ketorolac groups required rescue analgesic medication more frequently and/or larger supplemental dosages of fentanyl than the two opioid analgesic treatment groups. Compared with the three analgesic treatment groups, postoperative pain scores were only marginally higher in the control group. Ketorolac-treated patients had consistently (but not significantly) shorter recovery times to oral intake, ambulation, and discharge than those in the dezocine or fentanyl groups. No postoperative nausea, vomiting, or pruritis was reported in the ketorolac group. Conclusion: Compared with ketorolac 60 mg, fentanyl 100 μg and dezocine 6 mg produced a greater decrease in the propofol sedation requirement during MAC. However, the use of ketorolac in combination with propofol for MAC was associated with an improved recovery profile.