We assessed the effect of albumin infusion on weight loss and ventilation requirement in sick premature infants. Thirty infants, median gestational age 29 weeks, were entered into a randomised controlled trial, at a median of 2 days of age. The infants, all with an albumin level < or = 30 g/l, received either 5 ml/kg of 20% albumin or 5 ml/kg of their maintenance fluids (placebo), both given as part of the total daily fluid requirement. The response to the infusion was assessed by comparing two periods; 12 h immediately prior to the infusion and 12-24 h after the infusion. Albumin infusion was associated with a significant increase in albumin level and a significant reduction in weight, but in the placebo group there was a significant increase in weight. There were, however, no significant changes in the peak inspiratory pressure in response to either infusion. There was only a modest reduction (< 15%) in the inspired oxygen concentration, which occurred in both groups, but reached statistical significance only following the albumin infusion. We conclude that our results suggest that albumin infusion in "hypoalbuminaemic" sick preterm infants is unlikely to alter their respiratory status.