To examine the relationship of cigarette smoking, starting airway caliber, and airway pathology to nonspecific airway reactivity, we performed inhalation dose-response curves in 40 patients prior to lung surgery. Airway reactivity was assessed by the provocative concentration of methacholine (n = 30) or histamine (n = 10) resulting in a 20% fall in FEV1 (PC20). All patients had measurements of maximal expiratory flow, diffusing capacity, and lung volumes preoperatively. After resection, the specimens were graded for small and large airways pathology and emphysema. The patients were divided into 4 groups: PC20 less than or equal to 1 mg/ml (n = 6), PC20 1 mg/ml to less than or equal to 4 mg/ml (n = 11), PC20 4 mg/ml to less than or equal to 16 mg/ml (n = 14), and PC20 greater than 16 mg/ml (n = 9). Subjects with PC20 less than or equal to 1 mg/ml had reduced measurements of maximal expiratory flow (FEV1, FEV1/FVC, Vmax50, and Vmax25). There were no differences in measurements of cartilaginous airway disease among the 4 reactivity groups. We found significant independent correlations of cigarette consumption, membranous bronchiole inflammation, and maximal expiratory flow rates (FEV1 and Vmax50) to PC20. The data suggest that cigarette smoking, starting airway caliber, and airway inflammation are associated with nonspecific airway reactivity and that each exerts an effect on airway reactivity that is not dependent on the contribution of the other 2 factors.