It was the objective of this study to compare the efficacy of oral dexamethasone and inhaled budesonide in children hospitalized with croup, using a three-way, double blind, randomized, placebo-controlled clinical trial design. The trial was carried out in the Emergency Department Observation Ward of a tertiary pediatric hospital. The subjects for the study were 80 children (age range 5 to 158 months) who were hospitalized with croup. Children received either 2 mg of nebulised budesonide, dexamethasone syrup (0.6 mg/kg) or a placebo. Median duration of hospitalization was shorter for children treated with dexamethasone (12 hr) and budesonide (13 hr) compared to placebo (20 hr) (P < 0.03). There was no significant difference in hospitalization time between children treated with dexamethasone and budesonide. Median time to a croup score of < or = 1 was shorter for children treated with dexamethasone (2 hr) or budesonide (3 hr) compared to those who received placebo (8 hr) (P < 0.01). Croup scores for both steroid groups were significantly lower than the placebo group by 1 hr and remained so subsequently. The croup scores did not differ significantly in the 2 steroid treated groups. Six of the 30 children (20%) in the placebo group required adrenaline after the first hour compared to none of the 50 children in the steroid treated groups (P < 0.02). We conclude that oral dexamethasone and inhaled budesonide are both effective in reducing symptoms and duration of hospitalization in children with croup.