Inhaled corticosteroid therapy has proven efficacy for asthmatics, but the benefit for patients with chronic obstructive pulmonary disease (COPD) is less well supported. We hypothesized that withdrawal of inhaled steroids in elderly patients with severe irreversible airway obstruction would not lead to a deterioration in respiratory function. We designed a prospective, double-blind, randomized, placebo-controlled, crossover study to follow spirometry, quality of life questionnaire, six-minute (6-min) walk test, and sputum markers of inflammation during a 6-wk placebo treatment period and a 6-wk treatment period with beclomethasone dipropionate (BDP), 336 microg/d. There were 24 men receiving BDP who entered the study; 15 completed the study. Their mean age was 66.9 +/- 1.9 yr, and mean FEV(1) was 1.61 +/- 0.1 L (47% of predicted). There was a significant decrease in the mean FEV(1 )while using the placebo inhaler (1.70 L versus 1.60 L, baseline versus placebo: 95% CI, 0.002 to 0.195; p < 0.05). There was a decrease in the mean percentage change in FEV(1) for the study subjects during the placebo treatment period as compared with the BDP treatment period (-6.28 versus 5.03%, placebo versus BDP: 95% CI, -23.38 to 0.76; p = 0.06). Six-minute walk test results and sputum analysis for cell count and differential were not significantly different during placebo and BDP treatment periods. Borg scale assessment of dyspnea after exercise was increased while using the placebo inhaler as compared with baseline, and decreased during the BDP treatment period. Chronic Respiratory Disease Questionnaire (CRQ) scores revealed no significant difference between placebo and BDP. This study has demonstrated that in elderly patients with severe irreversible airway obstruction, withdrawal of inhaled corticosteroid therapy leads to a deterioration in ventilatory function and increased exercise-induced dyspnea.