A recent sevenfold increase in the annual incidence of chronic neonatal lung disease (CNLD) on an intensive care unit was attributed to the early administration of intravenous lipid (IVL) in ventilated preterm neonates. When logistic regression was used to eliminate other confounding variables, early delivery of IVL was independently associated with an eight-fold increase in the likelihood of CNLD. Consequently, we designed a prospective study to detect a halving of the incidence of CNLD by delaying IVL administration from 5 days (as is routine practice) to 14 days. Sixty-four parenterally fed preterm neonates weighing < 1,500 g at birth were randomised to receive IVL either on day 5 or day 14. Analysis was by intention to treat, since several infants in the latter group required no parenteral nutrition by day 14. Our results showed that the relative risk (95% confidence interval) of CNLD at 28 postnatal days was 1.15 (0.81-1.62); at 36 weeks postconception, it was 1.08 (0.59-1.99). A study population of > 2,000 would be required to determine whether these relative risks were significantly different from 1.