Background Serum levels of eosinophil cationic protein are an indirect measure of airway inflammation in asthma. It is proposed that the extent to which bronchoconstriction or airway inflammation contributes to airflow obstruction in acute asthma may determine responsiveness to bronchodilator therapy. Objective To test the hypothesis that subjects with acute asthma exacerbations who respond poorly to inhaled bronchodilator treatment may have more marked airway inflammation than those who respond well to identical therapy. Methods Forty-eight asthmatic children who visited the emergency room due to acute exacerbations were studied. Serum levels of eosinophil cationic protein were measured at the time of acute exacerbations and of clinical remissions at acute exacerbation, FEV 1 was assessed before and after the administration of aerosolized salbutamol. Results The mean serum level of eosinophil cationic protein at acute exacerbation (41.1 ± 12.8 microgram/L) was significantly higher ( P < .01) than that at clinical remission (30.0 ± 8.5 microgram/L) in the study population. The level at acute exacerbation was even higher in group A (n = 18: postbronchodilator FEV 1 < 75% predicted) than in group B (n = 30: postbronchodilator FEV 1 >/= 75% predicted), whereas both groups showed similar levels at clinical remission. The level at acute exacerbation correlated positively with severity of exacerbation (r = .47, P < .01) and negatively with bronchodilator responses (r = −.56, P < .01). This negative correlation was valid among subjects with a similar degree of exacerbation. Conclusion A higher level of eosinophil cationic protein at acute asthma exacerbation was associated not only with more severe exacerbation but also with a lower degree of bronchodilator responsiveness. This suggests that degree of airway inflammation may be one determinant of degree of responsiveness to initial bronchodilatory therapy at acute asthma exacerbation.