The question addressed in this study was the ability of young children to use a dry-powder inhaler, Turbuhaler(R). One hundred and sixty five children suspected of asthma, equally distributed in one year age-groups from 6 months to 8 yrs, inhaled from a Pulmicort Turbuhaler(R), 200 mu g budesonide.dose(-1), through a filter. The amount of drug trapped by the filter was used as a measure of drug released to the patient. None of the children had prior experience in the use of a Turbuhaler(R), but they were instructed carefully, together with their parents, in the clinic. The median dose released revealed an age-dependent increase, with a considerable scatter. Accordingly, the dose delivered could not be predicted in these young children. The limitation to effective use in young children appeared to be lack of sufficient co-operation, rather than physical limitations, as even some very young children appeared to obtain a sufficient activation of the device. It is likely that repeated training at home may improve these findings. In conclusion, our results indicate that dry-powder inhalers are not reliable in all circumstances for treatment of young children, and that careful and repeated tuition is required if such devices are to be used.