To evaluate comparative bronchodilator efficacy with chronic airflow obstruction (CAO), we randomly administered to ten patients week-long treatments consisting of: inhaled metaproterenol from a metered dose canister (1.30 mg six times a day) and doses of a sustained-release theophylline formulation sufficient to achieve plasma levels of 10-15 micrograms/ml; metaproterenol-placebo; theophylline-placebo; or placebo-placebo. At the end of each period, treatment responses were evaluated by spirometric tests, by exercise tolerance (12 minute walk and progressive cycle ergometry) and by subjective perception of dyspnea (oxygen cost diagram and breathlessness rating). Metaproterenol as a single treatment caused statistically significant improvements in spirometric variables and in the breathlessness rating. Theophylline as a single treatment caused significant changes in none of the test variables, though favorable trends were observed. Combined drug therapy was significantly better than metaproterenol only in the case of the breathlessness rating. We conclude that in the treatment of patients with CAO, inhaled metaproterenol is superior to oral theophylline. Our results permit no definite conclusion concerning added benefits of combined drug therapy.