Inhaled corticosteroids are often used in the treatment of stable chronic obstructive pulmonary disease (COPD), however, studies of these agents have had mixed results. Previous trials have often excluded subjects with bronchodilator response, have failed to evaluate effect on gas exchange, and have usually looked at only post- rather than prebronchodilator forced expiratory volume (FEV). Our objective was to better assess the efficacy of topical corticosteroids in the treatment of COPD. We used a prospective, randomized, double-blinded, placebo-controlled, crossover study at the Outpatient Department, Department of Veterans Affairs Medical Center. Thirty-six COPD patients with a mean (+/- SD) FEV1 of 1.10 +/- 0.43 L, with or without significant bronchodilator response participated in the study. Subjects received a 3-month course of inhaled fluticasone propionate (220 micro g/puff) or identical-appearing placebo by metered-dose inhaler at 2 puffs twice daily, followed by crossover to the alternative inhaler for an additional 3 months. Fluticasone treatment resulted in a higher prebronchodilator FEV1 (1.17 +/- 0.08 L [mean +/- SEM] versus 1.07 +/- 0.08 L, p = 0.001), a higher PaO2 (66.6 +/- 1.4 mmHg versus 63.6 +/- 1.6 mmHg, p = 0.002), and a better dyspnea score on the chronic respiratory questionnaire (3.70 +/- 0.18 versus 3.47 +/- 0.19, p = 0.03). A trend towards fewer exacerbations with fluticasone did not quite meet statistical significance (p = 0.11). Inhaled fluticasone over 3 months improved prebronchodilator airflow obstruction and oxygenation while decreasing dyspnea in moderate to severe COPD. Postbronchodilator FEV1 was not significantly changed.