Exposure of humans to ambient levels of ozone (O(3)) causes inflammatory changes within lung tissues. These changes have been reported for the "initial" (1- to 3-h) and "late" (18- to 20-h) postexposure periods. We hypothesized that at the late period, when protein and cellular markers of inflammation at the airway surface remain abnormal and the integrity of the epithelial barrier is compromised, bronchial reactivity would be increased. To test this, we measured airway responsiveness to cumulative doses of methacholine (MCh) aerosol in healthy subjects 19+/-1 h after a single exposure to O(3) (130 min at ambient levels between 120 and 240 parts/billion and alternate periods of rest and moderate exercise) or filtered air. Exposures were conducted at two temperatures: mild (22 degrees C) and moderate (30 degrees C). At the late period, bronchial reactivity to MCh increased, i.e., interpolated dose of MCh leading to a 50% fall in specific airway conductance (PC(50)) was less after O(3) than after filtered air. PC(50) for O(3) at 22 degrees C was 27 mg/ml (20% less than the PC(50) after filtered air), and for O(3) at 30 degrees C it was 19 mg/ml (70% less than the PC(50) after filtered air). The forced expiratory volume in 1 s (FEV(1)) at the late time point after O(3) was slightly but significantly reduced (2.3%) from the preexposure level. There was no relationship found between the functional changes observed early after exposure to O(3) and subsequent changes in bronchial reactivity or FEV(1) at the late time point. These results suggest that bronchial reactivity is significantly altered approximately 1 day after O(3); this injury may contribute to the respiratory morbidity that is observed 1-2 days after an episode of ambient air pollution.