We tested the hypothesis that maximal bronchoconstriction in humans in vivo is limited by the maximal contractility of airway smooth muscle by comparison of complete in vivo and in vitro dose-response curves to methacholine in 10 nonasthmatic subjects who were scheduled for thoracotomy because of malignancies. The provocative dose of methacholine that produced a 10 and 20% decrease of baseline FEV1 (PD10,20 FEV1) and the maximal fall in FEV1 (MFEV1) at the response plateau to inhaled methacholine were determined prior to surgery. Small airway smooth muscle preparations, obtained from the 10 resected lung tissue specimens, were examined in vitro to determine the sensitivity (-log EC50) and maximal isotonic shortening (Smax) to methacholine. In addition, the relaxation responses to the beta-agonist I-isoproterenol were measured. The degree of small airways disease (SAD) was examined histologically. Nine subjects showed a maximal response plateau to inhaled methacholine in vivo. The maximal fall in FEV1 at the plateau was 26 +/- 3%. All airway smooth muscle preparations (n = 30) contracted to methacholine (-log EC50, 5.94 +/- 0.09; Smax, 1320 +/- 219 micron) and relaxed to I-isoproterenol (-log EC50, 7.60 +/- 0.11; maximal relaxation [Rmax], 87 +/- 3%). No significant correlations were found between Smax or Rmax of the airway smooth muscle in vitro and the MFEV1 in vivo, and between -log EC50 for methacholine or I-isoproterenol in vitro and PD10 or PD20 FEV1 for methacholine in vivo. The severity of SAD was significantly correlated with the degree of baseline airflow limitation (p less than 0.05), but not with in vivo or in vitro responses to methacholine.(ABSTRACT TRUNCATED AT 250 WORDS)