BACKGROUND: Many preterm babies who survive, having had respiratory distress syndrome (RDS) or not, go on to develop chronic lung disease (CLD). This is probably due to persistence of inflammation in the lung. Corticosteroids have powerful anti-inflammatory effects and have been used to treat established CLD. However it is unclear whether any beneficial effects outweigh the adverse effects of these drugs. OBJECTIVES: To determine if late (> 3 weeks) postnatal corticosteroid treatment vs control (placebo or nothing) is of benefit in the treatment of chronic lung disease (CLD) in the preterm infant. SEARCH STRATEGY: Randomised controlled trials of postnatal corticosteroid therapy were sought from the Oxford Database of Perinatal Trials, the Cochrane Controlled Trials Register, Medline, hand searching paediatric and perinatal journals, examining previous review articles and information received from practising neonatologists. SELECTION CRITERIA: Randomised controlled trials of postnatal corticosteroid treatment initiated at > 3 weeks of age in preterm infants with CLD were selected for this review. DATA COLLECTION AND ANALYSIS: Data regarding clinical outcomes including mortality before discharge, failure to extubate, infection, hyperglycaemia, hypertension, severe retinopathy of prematurity (ROP), necrotizing enterocolitis (NEC), gastrointestinal bleeding, intestinal perforation, echodensities on ultrasound scan of brain, need for home oxygen, glycosuria, need for late rescue with dexamethasone, cerebral palsy in survivors and blindness in survivors including a study reporting 5 year follow-up were abstracted and analysed using Revman 4.1. MAIN RESULTS: Delayed steroid treatment had no effect on mortality. Beneficial effects of delayed steroid treatment included reductions in failure to extubate by 7 or 28 days, need for late rescue treatment with dexamethasone, chronic lung disease at 36 weeks, death or CLD at 36 wk, and discharge to home on oxygen therapy. There was no increase in risk of infection, necrotising enterocolitis, or gastrointestinal bleeding. Short-term adverse affects included hyperglycaemia, glycosuria and hypertension. There was an increase in severe retinopathy of prematurity, of borderline significance in survivors, but no significant increase in blindness. Although there were increases in long-term neurologic sequelae, including abnormal neurologic examination and cerebral palsy, those apparent trends were offset by a trend in the opposite direction due to an increased incidence of death before follow-up assessment. REVIEWER'S CONCLUSIONS: The benefits of late corticosteroid therapy may not outweigh actual or potential adverse effects. Although there continues to be concern about an increased incidence of adverse neurological outcomes in infants treated with postnatal steroids (see also review of Early postnatal corticosteroids), this review of postnatal corticosteroid treatment for CLD initiated after three weeks of age suggests that late or delayed therapy may not significantly increase the risk of adverse long-term neurodevelopmental outcomes. Nonetheless, corticosteroids should still be reserved for infants who cannot be weaned from mechanical ventilation. The dose of dexamethasone and the duration of any course of treatment should still be minimized.