Eucapnic voluntary hyperventilation (EVH) of dry gas is a physiologic bronchoprovocation challenge useful in the diagnosis of asthma. To determine the best parameter and threshold for diagnosis and the proper timing of postchallenge measurements, we reviewed 120 challenges, comparing the decrement from baseline in FVC, FEV1, mean forced expiratory flow during the middle half of the FVC (FEF25-75%), and peak expiratory flow rate (PEFR) each at 0, 5, 10, and 20 min postchallenge. After adjustment to a standard minute ventilation of 30 times the baseline FEV1 for 6 min, the mean response by 90 mild asthmatics differed from 30 normal subjects in all four parameters (p < 0.0001). In asthmatics, maximum decline from baseline (mean +/- SEM) was as follows: FVC, 12.1 +/- 1.2%; FEV1, 19.7 +/- 1.7%; FEF 25-75%, 33.5 +/- 2.5%; and PEFR, 29.0 +/- 1.9%. Normal subjects had a maximum fall as follows: FVC, 2.9 +/- 0.7%; FEV1, 3.8 +/- 0.7%; FEF25-75%, 11.8 +/- 2.0%; and PEFR, 11.5 +/- 1.0%. Based on comparison of receiver operator characteristic curves, FEV1 was more accurate than FEF25-75% and equivalent to FVC and PEFR. A threshold of 10% change or greater in FEV1 had a specificity of 90%, with a sensitivity of 63.3%. A threshold of 15% or greater had a specificity of 100%, with a sensitivity of 53.3%. The FEV1 fell by 10% or more in 55 of 90 asthmatics at 5 or 10 min after hyperventilation. Measurements at 0 or 20 min added two additional positive responses. We conclude that in the proper clinical setting, subjects whose FEV1 declines by 10% or more at 5 or 10 min after EVH should be diagnosed as having asthma.