The role of glucocorticoids in the treatment of chronic obstructive pulmonary disease (COPD) is controversial. We have previously described high numbers of neutrophils and high concentrations of the inflammatory cytokines interleukin-8 (IL-8) and tumor necrosis factor-alpha (TNF-alpha), and of the cell activation markers eosinophil cationic protein (ECP), eosinophil peroxidase (EPO), myeloperoxidase (MPO), and human neutrophil lipocalin (HNL) in COPD patients as compared with controls, and have postulated that the cytokines TNF-alpha and IL-8 play a role in propagating the inflammatory response in COPD. We have now studied the effects of inhaled and oral glucocorticoids on these inflammatory indices in induced sputum. Initially, we studied the effect of a 2-wk course of inhaled budesonide (800 mg twice daily for 2 wk) in 13 patients with severe COPD (mean FDV1: 35% predicted). There was no clinical benefit in either lung function or symptom scores, and no significant change in the inflammatory indices as measured by total and differential cell counts and concentrations of TNF-alpha eosinophil activation markers ECP and EPO, and neutrophil activation markers MPO and HNL. Because the lack of anti-inflammatory effect might have been due to poor drug delivery as a result of severe airflow limitation, we undertook a study examining the antiinflammatory effect of oral prednisolone (30 mg daily for 2 wk) in patients with COPD and undertook the same measurements in 10 patients with atopic asthma. Sputum eosinophil numbers, ECP, and EPO were significantly reduced in the asthmatic patients but were not modified in COPD. This confirms the clinical impression that inhaled steroids have little antiinflammatory effect, at least in the short term in this group of patients, and suggests that the inflammatory process in COPD is resistant to the antiinflammatory effect of glucocorticoids.