This prospective study was undertaken to evaluate the efficacy and safety of midazolam as a sedative agent in 14 critically ill patients without coma admitted to the intensive care unit at Nihon University Itabashi Hospital, Japan. Adequate sedation (sedation score [SS] 4-6) was induced with 0.058- 0.372 mg/ kg midazolam and maintained with a dose range of 0.03-0.4 mg/kg per h. Most burn or trauma patients required higher midazolam doses than patients with cardiovascular disease. The most frequent adverse events observed were hypotension, heart-rate fluctuation and electrocardiogram abnormalities. Pharmacokinetic analysis of the population suggested that lower drug clearance rates correlated with presence of complications. Plasma concentrations (EC50) of SS 5, estimated by logistic regression analysis, varied among patients (mean 194 ng/ml). Midazolam infusion achieved successful sedation in this critical care setting, but the optimum dose differed by patient and was influenced by the presence of complications.