Background: The forced expiratory volume in one second (FEV1) is the gold standard to assess airway obstruction but its sensitivity in children has been disputed. It is hypothesized that the diagnosis capacity of spirometry in detecting airway obstruction may improve by taking into account the shape of flow-volume loop (FVL). The hypothesis was that the upwards concavity should be more closely associated with the inhomogeneous lung emptying compared to FEV1. Methods: The analysis involved spirometric data prospectively obtained in children aged 6 to 11 years with asthma (n=99), or at risk for respiratory sequelae after premature birth (n=46). There were respectively 130 and 27 age matched healthy controls. FEV1 was measured and an index of the FVL curvature (β angle) was calculated using different instantaneous expiratory flows and forced vital capacity. Larger value of β indicates more upwards concavity and supposedly more heterogeneous lung emptying. Diagnostic values were assessed using area under the receiver-operating characteristic (ROC) curve (AUC). Results: With reference to healthy controls, FEV1-z-score was significantly smaller in asthma (p=0,009) or premature infants (p=0,022) and β significantly larger in asthma groups vs controls (p<0,001) but not in children born prematurely. The ROC curve analysis showed significantly better discrimination between asthma and control by β than FEV1 (AUC respectively 0,70 and 0,61). On the other hand, the ability to discriminate children born prematurely from controls was not significantly different between FEV1 and β (p=0,370). Conclusions: Compared to FEV1, β shows significantly better detection of airway obstruction in asthmatic but not in children born prematurely. It is surmised that the inhomogeneous lung emptying is thus detected as an indicator of early peripheral obstruction in asthmatic subjects. On the other hand, β in children born prematurely dont differ from control.