Although patients with chronic obstructive pulmonary disease (COPD) should stop smoking, some do not. In a double-blind, placebo-controlled study, we evaluated the effect of the inhaled glucocorticoid budesonide in patients with mild COPD who continued smoking. After a six-month run-in period, we randomly assigned 1277 subjects (mean age, 52 years; mean forced expiratory volume in one second [FEV1], 77 percent of the predicted value; 73 percent men) to twice-daily treatment with 400 microg of budesonide or placebo, inhaled from a dry-powder inhaler, for three years. Of the 1277 subjects, 912 (71 percent) completed the study. Among these subjects, the median decline in the FEV1 after the use of a bronchodilator over the three-year period was 140 ml in the budesonide group and 180 ml in the placebo group (P=0.05), or 4.3 percent and 5.3 percent of the predicted value, respectively. During the first six months of the study, the FEV1 improved at the rate of 17 ml per year in the budesonide group, as compared with a decline of 81 ml per year in the placebo group (P<0.001). From nine months to the end of treatment, the FEV1 declined at similar rates in the two groups (P=0.39). Ten percent of the subjects in the budesonide group and 4 percent of those in the placebo group had skin bruising (P<0.001). Newly diagnosed hypertension, bone fractures, postcapsular cataracts, myopathy, and diabetes occurred in less than 5 percent of the subjects, and the diagnoses were equally distributed between the groups. In patients with mild COPD who continue smoking, the use of inhaled budesonide is associated with a small one-time improvement in lung function but does not appreciably affect the long-term progressive decline.