OBJECTIVE To determine whether dexamethasone prevents postextubation airway obstruction in young children. METHODS Prospective, randomized, double-blind, placebo-controlled study. METHODS Pediatric intensive care unit of a university teaching hospital. METHODS Sixty-six children, < 5 yrs of age, intubated and mechanically ventilated for > 48 hrs. METHODS Patients were randomized to receive intravenous dexamethasone (0.5 mg/kg, maximum dose 10 mg) or saline, every 6 hrs for six doses, beginning 6 to 12 hrs before elective extubation. RESULTS Dependent variables included the presence of stridor, Croup Score, and pulsus paradoxus at 10 mins, 6, 12, and 24 hrs after extubation; need for aerosolized racemic epinephrine and reintubation. The dexamethasone and placebo groups were similar in age (median 3 months [range 1 to 57] vs. 4 months [range 1 to 59], p = .6), frequency of underlying airway anomalies (3/33 vs. 3/33, p = 1.0), and duration of mechanical ventilation (median 3.3 days [range 2.1 to 39] vs. 3.5 days [range 2.1 to 15], p = .7). The dexamethasone group had a lower frequency of stridor, Croup Score, and pulsus paradoxus measurement at 10 mins and at 6 and 12 hrs after extubation. Fewer dexamethasone-treated patients required epinephrine aerosol (4/31 vs. 22/32, p < .0001) and reintubation (0/31 vs. 7/32, p < .01). Three patients exited the study early-one patient in the dexamethasone group had occult gastrointestinal hemorrhage and one patient in each group had hypertension. CONCLUSIONS Pretreatment with dexamethasone decreases the frequency of postextubation airway obstruction in children.