We sought to examine the effect of the introduction of dexamethasone therapy on health, growth, and neurodevelopmental outcome in very low birth weight (VLBW) infants at 20 months of age. We compared outcomes in all 86 VLBW infants (mean birth weight 871 gm, mean gestational age 26.4 weeks) who were ventilator dependent on day 21 of life during the 2 years preceding October 1988 (period 1), when dexamethasone therapy became accepted clinical practice in our unit, with outcomes in all 124 infants (mean birth weight 891 gm, mean gestational age 26.9 weeks) with similar ventilator status during the subsequent 2 years (period 2). In addition, we compared outcomes in infants who received dexamethasone during period 2 with those in a concurrent cohort of less ill infants who were not given dexamethasone. There were no significant differences between periods 1 and 2 in mortality rates after 21 days (17% vs 21%), need for home oxygen (23% vs 25%), oxygen dependence at 20 months of corrected age (11% vs 10%), rate of neurosensory impairment (24% vs 25%), and mean Bayley Mental scores (81.5 vs 77.2) or Psychomotor Development Index (81.6 vs 71.1). Infants who received dexamethasone during period 2 had significantly more severe lung disease and poorer respiratory, growth, and developmental outcomes. We conclude that VLBW infants with ventilator-dependent chronic lung disease have very poor outcomes, even when treated with dexamethasone. More information is needed from prospective, randomized trials before dexamethasone can be accepted as routine therapy for chronic lung disease.