The role of budesonide in controlling chest symptoms in infants was assessed. It was administered from a metered dose inhaler into a large volume spacer (Nebuhaler) with attached Laerdal mask. Twenty nine infants were recruited into a double blind crossover trial. Five defaulted. The remaining 24 (mean age 11 months) were assessed for the tolerance of the device, adverse effects, treatment failures, parental opinion, and daily symptom and treatment records. Twenty tolerated the modified Nebuhaler. One developed meningitis on placebo. Two experienced exacerbations on placebo. Eleven of 18 parents whose children completed the maximum treatment preferred budesonide to placebo and one preferred placebo. Mean symptom scores on budesonide were better than on placebo for the 15 children with complete symptom records. Fewer bronchodilator doses were used while taking budesonide. Our findings indicate that budesonide given in this way is an effective treatment for infants who may need prophylaxis for their wheezing.