This is the first randomized, controlled trial to examine the effects of treatment with inhaled steroid on cough discomfort as well as inflammatory indexes. Budesonide 800 |ig daily for two or four weeks had no effect. The results emphasize the importance of inclusion of a placebo arm to the treatment because any apparent improvements seen with bu-desonide were matched by placebo. The lack of beneficial effect of budesonide on cough scores or sputum inflammatory indexes is probably because the cause of the cough was not associated with sputum eosinophilia. Sputum eosino-philia seems to predict benefit from steroid treatment in chronic cough without asthma , in asthma and in smokers with chronic airflow limitation . Our results are confirmed by Gibson et al , who also reported that patients with nonasthmatic chronic cough who did not respond to treatment with inhaled steroid were those without eosino-philia in bronchoalveolar lavage.