The fraction of exhaled nitric oxide (FENO) is reduced by anti-inflammatory treatment in asthma. However, the FENO level is also regulated by individual demographics and there is considerable variation among clinically stable patients.We hypothesized that some demographics may be responsible for persistent FENO elevation despite inhaled corticosteroids (ICS) therapy in asthma.This was a prospective observational study. We initially screened 250 stable asthmatics and determined the FENO cut-off point for identifying poorly controlled asthma defined by one of the following criteria: Asthma control test20, or forced expiratory volume in one-second % of predicted80%, or peak expiratory flow variability80% (Study 1). After 12-weeks, 229 patients who maintained high or low FENO were selected and the independent factors which might contribute to a high FENO were examined (Study 2).A FENO level39.5 p.p.b. yielded 67% sensitivity and 76% specificity for identifying the patients with poorly controlled asthma. The persistent high FENO group (≥ 40 p.p.b.) was more likely to be ex-smokers, to show evidence of atopy (positive specific IgE, higher serum IgE and blood eosinophils), and to have allergic comorbidities. Especially, past smoking history, blood eosinophils, and chronic rhinosinusitis were identified to be independent predictors of high FENO. Neither the dose of ICS nor other medication use showed any difference between the groups.These results suggested that past smoking history, blood eosinophilia, and chronic rhinosinusitis are involved in the persistent airway inflammation detected by FENO. Although their relative contributions on FENO values should be further quantified, clarification of the features of the subjects with high FENO might provide clues for adjustment of the treatment approach in asthma.