Study Objective: To investigate the adequate infusion dose regimen of midazolam to induce sedation with the Ramsay score 4 with rapid onset during combined spinal and epidural anesthesia Design: Prospective, randomized study. Setting: Operating room of a university hospital. Patients: 80 ASA physical status I and II patients aged 30 to 70 years, undergoing combined spinal and epidural anesthesia. Interventions: Patients were randomized to four groups of 20 patients each at random. After starting surgery, an infusion of midazolam 0.3, 0.6, 0.9, or 1.2 mg/kg/hr was started. When patients closed their eyes spontaneously, the infusion dose was decreased to one half of the initial dose. At 2.5 and 5 minutes after decrease the dose and at 5-minute intervals for the first 30 minutes then at 15-minute intervals thereafter until the end of surgery, infusion dose was adjusted by decreasing to one half or increasing to twice to keep the Ramsay score 4. Measurements and Main Results: The number of patients who required oxygen was significantly larger in the groups received 0.9 and 1.2 mg/kg/hr. Eleven patients with 0.9 mg/kg/hr and 17 patients with 1.2 mg/kg/hr at 5 minutes, but no patients with 0.3 or 0.6 mg/kg/hr showed Ramsay score 6 (heavy sedation). Amnesia was observed in all patients. Time to eye closure was dose dependently faster with the larger doses. Conclusions: During combined spinal and epidural anesthesia, midazolam 0.6 mg/kg/hr given until closing of the eyes (for 1.6 min) followed by midazolam 0.75 mg/kg/hr provides rapidly induced sedation, with a Ramsay score of 4 and amnesia with stable hemodynamics and respiration. (C) 2004 by Elsevier Inc.