OBJECTIVE: To determine whether fentanyl infusions given to premature infants with respiratory distress syndrome reduce stress and improve long- and short-term outcome. METHODS: Twenty premature infants undergoing mechanical ventilation for respiratory distress syndrome were randomly assigned, in a double-blind fashion, to receive fentanyl by continuous infusion or a volume-matched placebo infusion. A behavioral state score was used to assess the infants' behavior. Cortisol and 11-deoxycortisol levels were measured as physiologic markers of stress. Urinary 3-methyl histidine/creatinine molar ratio was determined and the fractional excretion of urea was measured to assess catabolic state. Ventilatory indexes were recorded for each infant. RESULTS: Infants receiving fentanyl showed significantly lower behavioral state scores (p <0.04) and lower heart rates (p <0.001) than those receiving placebo. 11-Deoxycortisol levels were lower in the fentanyl group on days 3, 4, and 5 of the study (p <0.003). 3-Methyl histidine/creatinine ratios and fractional excretion of urea were not significantly different between the two groups. On the third day of the study, infants receiving fentanyl required a higher ventilator rate (p <0.01), higher peak inspiratory pressures (p <0.001), and higher positive end-expiratory pressure (p <0.0001) than those receiving placebo. There was no difference in long-term outcome with respect to the incidence of bronchopulmonary dysplasia, intraventricular hemorrhage, or sepsis or with respect to the duration of ventilator use. CONCLUSIONS: Although there was a reduction in stress markers in the infants receiving fentanyl, we were unable to demonstrate an improvement in catabolic state or long-term outcome. In addition, the infants receiving fentanyl required higher ventilatory support in the early phase of respiratory distress syndrome than did those receiving placebo. (J PEDIATR 1996;129:140-5)