Excessive inflammation seems important in chronic obstructive pulmonary disease (COPD), particularly during exacerbations of the disease. Exhaled nitric oxide concentration ([NOexh]) is a sensitive marker of bronchial inflammation in asthma; it is unclear if this is also the case in COPD. This study: 1) quantifies [NOexh] in patients with COPD (during an exacerbation and while clinically stable); 2) investigates the response of [NOexh] to i.v. steroid therapy, and its potential relationship with other relevant physiological variables; and 3) assesses the relative contributions of the central and peripheral airways to [NOexh] by collecting exhaled air in two different bags connected in series. Seventeen COPD patients (forced expiratory volume in one second (FEV1) 37.6+/-3.4% of the predicted value (+/-SEM)) hospitalized because of an exacerbation of the disease (arterial oxygen tension (Pa,O2) (7.46+/-0.72 kPa 56.1+/-5.4 mmHg), arterial carbon dioxide tension (Pa,CO2) 5.63+/-0.37 kPa 42.3+/-2.8 mmHg), pH 7.41+/-0.02) and 10 healthy subjects that served as controls were studied. On admission, [NOexh] in COPD was higher than normal (41.0+/-5.1 versus 13.3+/-0.8 parts per billion (ppb), respectively, p<0.001). Despite i.v. steroid therapy, [NOexh] remained elevated throughout recovery (37.9+/-4.8 ppb, p<0.001) until discharge (40.9+/-4.3 ppb, p<0.001). In contrast, when the patients were clinically stable (several months later), [NOexh] was significantly reduced (15.8+/-3.8 ppb, p<0.001), and no longer different from control values. [NOexh] was not related to any of the physiological variables measured during recovery (pulmonary gas exchange) or at discharge (forced spirometry, lung volumes, diffusing capacity). Finally, the contribution of the central and peripheral airways to [NOexh] was not different at any point in time. These results indicate that during exacerbations of chronic obstructive pulmonary disease, the exhaled nitric oxide concentration: 1) is higher than normal; 2) is not reduced acutely by i.v. steroids but is normalized several months after discharge; 3) is unrelated to several physiological indices of disease severity; and 4) appears to be produced homogeneously in central and peripheral airways. Overall, these results are different from those reported in asthma, suggesting that different inflammatory mechanisms are operating in both diseases.