The purpose of this study was to examine airway responsiveness, sputum cells and the effects of inhaled corticosteroid in the chronic cough syndrome associated with eosinophilic bronchitis. We studied nine consecutive referrals with chronic cough, sputum with >10% eosinophils, normal spirometry, and normal methacholine airway responsiveness. Clinical assessment, sputum analysis, allergy skin tests and a methacholine inhalation test were performed at the first visit. Peak expiratory flow (PEF) was measured twice daily for 1 week followed by an adenosine monophosphate (AMP) inhalation test. Subjects were then treated with inhaled beclomethasone 0.4 mg twice daily for 7 days. Sputum analysis and measurement of methacholine responsiveness were then repeated. Excessive airway narrowing to methacholine was not present in any of the subjects. A methacholine plateau response was present in five subjects. Hyperresponsiveness to AMP was absent in six of the nine subjects, and PEF variability was not increased for eight subjects. Corticosteroid therapy led to a reduction in sputum eosinophil counts from 40.1 (so 21.4)% to 4.0 (4.5)% but there was no significant change in metachromatic cell counts (0.8 so 0.5% vs 0.6 sd 0.6%) or total cell counts. Methacholine responsiveness improved within the normal range in the three subjects in whom it could be determined. Chronic cough associated with eosinophilic airway inflammation can occur in the absence of variable airflow obstruction (asthma) and can improve after treatment with inhaled corticosteroid. This treatment can reduce the level of methacholine responsiveness within the normal range and reduces sputum eosinophils but not mast cells. These results suggest that the occurrence of variable airflow obstruction depends on the baseline level of methacholine responsiveness, the degree of eosinophilic infiltration and the degree to which methacholine responsiveness becomes heightened.