OBJECTIVE:The aim of this study was to evaluate the pharmacodynamics and safety of midazolam after intravenous infusion or oral administration in preterm infants. METHODS ]PATIENTS WERE RANDOMLY ASSIGNED TO INITIALLY RECEIVE MIDAZOLAM 0.1 MG/KG AS A 30-MINUTE INTRAVENOUS INFUSION OR AN ORAL BOLUS DOSE. IF PATIENTS STILL MET THE INCLUSION CRITERIA, THEY THEN RECEIVED MIDAZOLAM VIA THE ALTERNATE ROUTE (AFTER AN INTERVAL OF ≥72 HOURS). PHARMACODYNAMIC MEASUREMENTS CONSISTED OF A COMFORT® SCORE (A PREVIOUSLY VALIDATED SEDATION SCALE FOR PAEDIATRIC PATIENTS) AT BASELINE AND AT 0.5, 1, 2, 4 AND 6 HOURS POSTDOSE. MIDAZOLAM AND 1-OH-MIDAZOLAM CONCENTRATIONS WERE MEASURED AND VITAL SIGNS WERE RECORDED AT ALL PHARMACODYNAMIC MEASUREMENT TIMEPOINTS: RESULTS:A total of 24 infants were enrolled of whom seven received both intravenous and oral midazolam, 13 received only intravenous midazolam, and four received only oral midazolam. Overall, mean COMFORT® scores decreased (i.e. sedation increased) significantly within 30 minutes after intravenous (p S 0.05) and within 1 hour after oral (p = 0.003) midazolam administration. In 45% of patients the COMFORT® scores decreased little or not at all after midazolam, which was similar after both oral and intravenous administration. The sedative response to midazolam did not differ after intravenous or oral administration. No relationship was found between overall COMFORT® scores or change in COMFORT® score from baseline and midazolam, 1-OH-midazolam, or midazolam plus 1-OH-midazolam concentrations. Diastolic blood pressure decreased significantly after intravenous (approximately 11%) but not after oral midazolam administration. No serious adverse events were reported. CONCLUSIONS:Midazolam administered as a 30-minute intravenous infusion or oral bolus dose appears to be effective and well tolerated in a small majority of preterm infants. However, a considerable number of neonates do not appear to respond to midazolam. The lack of response may be due to the fact that patients truly experienced therapeutic failure and/or consequent to the inability of the COMFORT® score to adequately reflect sedation uniformly in sick preterm infants.