Ten children with severe asthma, who had been well controlled on maintenance betamethasone valerate aerosol for an average of 11 months, were given placebo aerosols without their knowledge. The period of placebo substitution was campared with one 28-day period of betamethasone valerate therapy beforehand, and two 28-day periods afterwards. Symptoms were increased during the placebo period, and patients did not return to their previous well-controlled state until the second month after reinstitution of therapy. Changes in the means of twice-daily peak expiratory flow readings (PEFR) followed the same pattern as changes in symptoms. The exacerbation of asthma which occurred during placebo treatment was accompanied by a widening in the diurnal variation between morning and evening PEFR. In comparison with the previous period, morning PEFR fell by a greater amount than evening PEFR. Standardized running tests suggest an increase in exercise-induced bronchoconstriction and in the Exercise Lability Index when the child was receiving only placebo treatment as compared with betamethasone valerate treatment. The trial provided further evidence of the efficacy of betamethasone valerate aerosol in the prophylatic therapy of severe childhood asthma. As 2 of these children were able to discontinue long-term therapy it is unlikely that this drug causes dependency.