The pharmacokinetics of furosemide were studied longitudinally during long-term administration in 10 very low birth weight infants with bronchopulmonary dysplasia. Mean birth weight of the infants was 829 +/- 217 g, mean gestational age at birth was 26.6 +/- 2.9 weeks, and mean postnatal age at the start of therapy was 2.4 +/- 1.0 weeks. Serial determinations of furosemide pharmacokinetic parameters were performed during 2 weeks to 3 months of long-term therapy. Plasma half-life was prolonged in infants less than 31 weeks postconceptional age (gestational + postnatal age), frequently exceeding 24 hours. All infants less than 29 weeks postconceptional age whose dosing schedule was once every 12 hours accumulated furosemide to potentially ototoxic levels. Furosemide renal clearance increased and plasma half-life decreased in association with increasing postconceptional age. Furosemide secretory clearance was very low in patients less than 31 weeks postconceptional age, resulting in a reliance on glomerular filtration to deliver drug to its main site of action within the lumen of the loop of Henle. Thus elevated plasma levels may be required to ensure adequate luminal delivery and adequate diuresis in these infants with low secretory clearance. Nevertheless, the current dosing schedule (once every 12 hours) of furosemide should be modified to once every 24 hours in infants of low postconceptional age to avoid possible toxic effects.