BACKGROUND Forced expiratory flow-volume curves are commonly used to assess the degree of airflow obstruction in patients with chronic obstructive pulmonary disease (COPD). In mechanically ventilated subjects, expiratory airways obstruction can only be estimated from relaxed expirations. The aim of this study was to quantify the degree of airways obstruction from relaxed expiratory flow-volume curves in mechanically ventilated patients with COPD. METHODS As measure of airflow obstruction, the effective time constant during the last 50% of expired volume (tau) was calculated. For bedside monitoring, tau was recalculated as the slope of the flow during the last 50% of expired volume (SF50). In order to study reproducibility, the variables were calculated from consecutive breaths and at different levels of end-expiratory lung volume (EEV). The SF50 and the tau-were correlated with the forced expiratory volume in 1 s (FEV1) measured prior to the start of ventilatory support. RESULTS Twenty-seven patients were studied with a FEV1 expressed as percentage predicted of 31 +/- 12% (mean +/- SD). The SF50 amounted to 19 +/- 10 degrees. A positive linear correlation was established between SF50 and the FEV1, (%pred), (r = 0.90, P < 0.0001). The tau showed an exponential relationship with FEV1 (%pred), (r2 = 0.78). From 5 consecutive breaths the mean variation coefficient of SF50 was 5 +/- 2%. Changes of delta EEV from 0.05 to 1.00 L did not affect the SF50-values. In 12 patients, mechanically ventilated for respiratory diseases other than COPD, mean tau and SF50 were significantly different from the COPD-patients (P < 0.0001). CONCLUSIONS This study indicates that relaxed expiratory flow-volume curves can be used to assess airflow obstruction in mechanically ventilated patients with COPD. This information can be used to adapt ventilatory settings.