We have studied prospectively, in nine children requiring sedation to facilitate mechanical ventilation, the metabolic, biochemical and haemodynamic effects of infusion of propofol. Children were given infusions of propofol 1-4mg kg-1 h-1 and fentanyl 1-5 micrograms kg-1 h-1 for 48 h. Heart rate, arterial pressure, central venous pressure, fluid balance and urine output were recorded hourly and sedation scores every 4 h. In addition to routine haemodynamic and biochemical measurements in the intensive care, 6-hourly arterial blood-gas analysis and 12-hourly measurements of serum concentrations of glucose, lactate and electrolytes, renal function, triglycerides and liver function tests were performed. Urine was analysed for ketones. There were no significant differences in haemodynamic or biochemical variables during the 48-h period. In this small sample of children, propofol combined with fentanyl provided excellent sedation with no evidence of cardiac, renal or hepatic impairment. Under these very proscriptive conditions we did not encounter lipaemia or acidosis with infusion of propofol. Thus propofol may be a safe sedative agent for use in paediatric intensive care if used appropriately. Further large scale studies are needed to determine if warnings against the use of this agent in paediatric intensive care units are justified.