Background : When appropriate premedication is required for pediatric patients, the route of drug administration and the patient's age may affect the drug response. This study was designed to evaluate the premedicative effects of intranasal, oral, and rectal midazolam in preschool(1∼6 year) and school (6.1∼10 year) ages. Methods : One hundred fourteen children aged 1∼10 years were randomly allocated into three groups to receive midazolam via intranasal (0.3 mg/kg), oral (1.0 mg/kg), or rectal (1.0 mg/kg) route. Sedation scores were evaluated at the arrival in preanesthetic room, drug administration, 5, 10, 20 and 30 min after drug administration, separation from parent, mask application, and induction with inhalational agent. Time to sedation scores of 3 and 4 and time to complete recovery from general anesthesia were recorded. Results: At the drug administration, the incidence of crying was significantly higher in nasal group than in oral and rectal groups, especially in pre-school age group (87.5, 23.5. 40.9% for nasal, oral and rectal groups, respectively). At 5, 10 and 20 min after drug administration, sedation scores were significantly higher in nasal and rectal groups than in oral group. At separation, mask application and inhalational induction, sedation scores were significantly higher in oral and rectal groups than in nasal group. Time to sedation score of 3 and time to complete recovery were significantly longer in oral group than in nasal and rectal groups. Conclusions: In pre-school age, almost all the children cried at drug administration in nasal group, and onset and recovery were prolonged in oral group, so rectal route was suitable. In school age, nasal route was appropriate because of the lower frequency of crying at the drug administration and rapid onset and recovery. In overall age, rectal route was better because of the lower frequency of crying and rapid onset and recovery. This study suggests that administration route should be considered according to the age of pediatric patient to obtain appropriate premedication for pediatric patients. (Korean J Anesthesiol 1998; 34: 730∼738)