Increased airway hyperresponsiveness is thought to be one of the phenomena underlying nocturnal airway obstruction in asthma. To investigate the mechanisms that influence and modulate this phenomenon, we compared circadian variations in airway responsiveness with AMP and propranolol with the circadian variation in airway responsiveness to methacholine. Inhalation provocation tests were performed at 16.00 and 04.00 h in 16 nonsmoking atopic asthmatic subjects (18 to 42 yr of age), prospectively assigned to Group 1 (mean circadian peak expiratory flow rate [PEFR] variation > or = 15%) and Group 2 (< 15%). The circadian change in airway responsiveness to AMP, in contrast to methacholine, was significantly related to the circadian PEFR-variation of the 16 subjects (r = 0.81, p < 0.001). In Group 1 (n = 7) geometric mean PC20 AMP decreased more than PC20 methacholine during the night (2.3 and 0.9 doubling concentrations respectively, p < 0.05), whereas no difference in baseline FEV1 was found at the same time points during the different study days. Geometric mean PC20 propranolol did not change during the night. Daytime PC20 propranolol and PC20 AMP, in contrast to PC20 methacholine, were significantly lower in Group 1 as compared with Group 2. Together, the results show a higher susceptibility to stimulation of "indirect" airway responsiveness in the subjects with increased circadian PEFR amplitude. This suggests that mast cell activation rather than primary changes in smooth muscle cell contraction may play a role in the development of nocturnal airway obstruction.