To assess the effects of theophylline in chronic obstructive pulmonary disease, we conducted a randomized, placebo-controlled, double-blind, crossover trial in 60 patients with severe but stable disease. The patients (mean age, 61 years) were studied before and after two months of placebo and two months of treatment with a sustained-release preparation of theophylline (10 mg per kilogram of body weight per day), administered orally. The two treatments were administered in a random order and separated by an eight-day washout period. After taking theophylline for two months (mean plasma concentration, 14.8 mg per liter), as compared with the two months of placebo, the patients had significant improvements in dyspnea, pulmonary gas exchange (partial pressure of arterial oxygen, 66 vs. 61 mm Hg [P less than 0.0001]; partial pressure of arterial carbon dioxide, 44 vs. 49 mm Hg [P less than 0.0001]), vital capacity (63 percent vs. 58 percent of the predicted value [P less than 0.0001]), and forced expiratory volume in one second (36 percent vs. 32 percent of the predicted value [P less than 0.0001]), with no significant change in airway resistance or functional residual capacity. Minute ventilation increased by a mean of 18 percent (P less than 0.0001) in the patients taking theophylline because of increased tidal volume, with no change in respiratory frequency. The respiratory-muscle performance of the patients taking theophylline improved by approximately 29 percent (P less than 0.0001), as indicated by a decline in the ratio of inspiratory pleural pressure during quiet breathing to maximal pleural pressure. We conclude that theophylline improves respiratory function and dyspnea in patients with severe chronic obstructive pulmonary disease and that these improvements are probably due to better respiratory-muscle performance.