Abstract Airway hyperresponsiveness is easily assessed by measuring the concentration or dose of an inhaled agonist that produces a defined response, e.g., PC20 or PD20. However, this measure does not assess excessive bronchoconstriction. We report the results of analyzing bronchial dose-response curves by measuring percent fall in vital capacity (delta FVC%) as the response rather than the PC20. In our analysis, delta FVC% was measured at the PC20, and therefore it was the dependent variable, whereas the concentration of agonist was the independent variable, in contrast to the usual bronchoprovocation tests in which the response is the independent variable and the dose is the dependent variable. We reasoned that a dose-dependent increase in gas trapping with histamine would detect excessive bronchoconstriction as a decrease in FVC; in contrast, PC20 measures only the ease of bronchoconstriction. In 10 patients with mild asthma the reproducibility of delta FVC% when FEV1 fell by 20%, i.e., at the PC20 concentration of histamine, taken from a greater than 6-s FVC on an otherwise standard histamine challenge test was comparable to that of PC20. In 10 healthy asymptomatic subjects there were only trivial falls (0.3%) in FVC to as much as 16 mg/ml histamine. In a retrospective study of 146 patients with mild asthma, the delta FVC% was normally distributed (13.2 +/- 5.5 SD%) and did not correlate with the number of beta 2-agonist prescriptions or the PC20, but it did correlate with the number of prescriptions written per month for oral prednisone (r = 0.55, p < 0.02). We conclude that delta FVC% when FEV1 falls by 20% is a safe method of detecting excessive bronchoconstriction, and it reveals that different asthmatics react in fundamentally different ways to the same agonist. This may be useful in detecting the asthmatic at risk for serious disease.