We examined the volume history effect on maximal expiratory flow (Vmax) in patients with asthma by comparing Vmax at 60% of VC from forced exhalation maneuvers begun just above FRC (partial, P) with those begun from TLC (maximal, M) and expressed the results as M-P ratios. In a clinic population with varying degrees of obstruction, we found that M-P ratios varied inversely with the severity of the obstructive process, i.e., the more severely obstructed patients had a fall in Vmax after a deep inhalation (DI). By contrast, equally severe obstruction acutely induced in subjects with mild asthma was associated with an increase in Vmax after a DI. There was no difference between the spontaneous versus the induced groups in degree of nonhomogeneity as assessed by single-breath nitrogen tests. Changes in specific airway conductance after a DI during spontaneous obstruction were in the same direction as the changes in Vmax, and the time course for restitution of airway caliber in subjects who showed bronchoconstriction after a DI was significantly longer than for those subjects who showed bronchodilatation. We conclude that in severe bronchial asthma with spontaneous obstruction, a DI produces an increase in severity that is opposite to the results found in acutely induced obstruction, and the time course for reestablishing baseline airway caliber is more prolonged. We suggest that mechanisms for and sites of obstruction vary between the 2 groups.