A proportion of patients with cystic fibrosis (CF) suffer from increased airway hyperreactivity but their response to bronchodilators is variable. Adrenergic agents may produce an increase, no change or a decrease in forced expiratory volume in 1 second (FEV1). We hypothesized that the variable response might be related to poor aerosol distribution caused by the presence of secretions. Therefore, in 11 children with CF and airway hyperreactivity the influence of pretreatment with either 0.9% saline, salbutamol, or ipratropium bromide on the methacholine challenge test was evaluated in a double-blind, randomized, cross-over study. FEV1 (mean +/- S.E.) did not change following pretreatment with saline, salbutamol, or ipratropium (1.64 +/- 0.22, 1.63 +/- 0.16 and 1.67 +/- 0.19, respectively). All patients demonstrated airway hyperreactivity with a PC20 below 8 mg/mL (geometric mean, 0.41 mg/mL) after saline pretreatment. Salbutamol inhalation significantly increased the PC20 to 1.24 mg/mL (P less than 0.01), but ipratropium bromide was found to be even more effective than salbutamol (PC20 = 7.37 mg/mL) (P less than 0.0001). We conclude that the variable response to bronchodilator is not secondary to impaired aerosol distribution since ipratropium bromide effectively blocked the response to methacholine. The improvement in PC20 without a change in baseline FEV1 following salbutamol suggests that the adrenergic agent altered the contractile mechanism of smooth muscle.