A defect in the smooth muscle of airways has been discarded as a possible cause of asthma in recent years because of the lack of correlation between airflow obstruction in patients and the contractile responsiveness of the isolated airway smooth muscle. Howard Mitchell and Malcolm Sparrow question the relevance of comparing parameters obtained from pharmacological dose-response curves (e.g. EC50) of strips of airways in vitro with those describing airways narrowing in vivo (e.g. resistance). They point out that in small airways the upper half of the dose-response curves seen in strips of airway wall is not represented in perfused tubular airway segments because they are fully constricted at or near the EC50 of the strip.