Although FEV1 improvement is routinely used to define bronchodilator (BD) response, it correlates poorly with clinical effects. Changes in lung volumes (LV) have shown better correlation with exercise tolerance and might be more sensitive to detect BD effects. We assessed the additional contribution of measuring LV before and after BD to detect acute improvement in lung function not demonstrated by FEV1, and the influence of the response criteria selected on this contribution. We analyzed 98 spirometries and plethismographies performed pre and post BD in patients with airflow obstruction (FEV1/FVC < 70%). BD response was defined for FEV1 and FVC as per ATS guidelines and for other LV as delta > or = 10% of baseline (delta > or = 5 and > or = 15% were also analyzed). FEV1 identified as responders 32% of patients. Greater proportions were uncovered by slow vital capacity (51%, p < 0.001), inspiratory capacity (43%, p < 0.05) and residual volume (54%, p < 0.001). Slow spirometry identified 11% of responders additional to those detected by FEV1 and FVC. Plethismography added 9% more. The magnitude of volume responses correlated with the degree of baseline hyperinflation. Percentages of responders varied greatly using different thresholds (delta > or = 5 and > or = 15%). Mean change and proportions of responders for each LV varied significantly (p < 0.05) whether change was expressed as percent of baseline or predicted values. A considerable proportion of patients with airflow obstruction shows acute response to bronchodilators identified by changes in lung volumes but not detected by an improvement in FEV1. The selection of LV response criteria has important influence on the magnitude of this additional detection.