Current assessment of COPD relies extensively on the use of spirometry, an effort-dependent maneuver. Impulse oscillometry (IOS) is a non-volitional way to measure respiratory system mechanics, but its relationship to structural and functional measurements in large groups of patients with COPD is not clear. We evaluated the ability of IOS to detect and stage COPD severity in the prospective ECLIPSE cohort of COPD patients defined spirometrically, and contrasted with smoking and non-smoking healthy subjects. Additionally, we assessed whether IOS relates to extent of CT-defined emphysema. We measured lung impedance with IOS in healthy non-smokers (n = 233), healthy former smokers (n = 322) or patients with COPD (n = 2054) and related these parameters with spirometry and areas of low attenuation in lung CT. In healthy control subjects, IOS demonstrated good repeatability over 3 months. In the COPD group, respiratory system impedance was worse compared with controls as was frequency dependence of resistance, which related to GOLD stage. However, 29-86% of the COPD subjects had values that fell within the 90% confidence interval of several parameters of the healthy non-smokers. Although mean values for impedance parameters and CT indices worsened as GOLD severity increased, actual correlations between them were poor (r ≤ 0.16). IOS can be reliably used in large cohorts of subjects to assess respiratory system impedance. Cross-sectional data suggest that it may have limited usefulness in evaluating the degree of pathologic disease, whereas its role in assessing disease progression in COPD currently remains undefined. Copyright © 2011. Published by Elsevier Ltd.