As part of a clinical trial of early intervention in chronic obstructive pulmonary disease (COPD) (the Lung Health Study), 5,733 smokers with mild to moderate airflow obstruction underwent methacholine challenge tests at baseline. All participants were randomized to receive either usual care (no intervention) or special intervention, consisting of intensive smoking cessation counseling and the prescription of a metered-dose inhaler containing either ipratropium bromide or placebo (two inhalations three times daily). For this report, we analyzed the relationship between baseline methacholine reactivity and subsequent change in lung function. Methacholine reactivity was expressed as a logarithmic function of the two-point slope of percent decline in FEV1 over the concentration of methacholine (LMCR). Using a random effects linear model, LMCR was found to be a strong predictor of change in FEV1% predicted, after controlling for baseline lung function, age, sex, baseline smoking history, and changes in smoking status. Significant interactions were found between reactivity and smoking behavior. In the first year, participants who quit smoking showed improvement in FEV1, whereas continuing smokers showed worsening, and between Years 1 and 5, lung function declined to a greater extent in continuing smokers than in sustained quitters. For both time periods, these quitter/smoker differences increased as a function of airway reactivity. These findings indicate that methacholine reactivity is an important predictor of progression of airway obstruction in continuing smokers with early COPD, independent of the baseline level of obstruct.