Idiopathic chronic eosinophilic pneumonia (CEP) is a rare disorder of unknown cause with nonspecific respiratory and systemic symptoms but rather characteristic peripheral alveolar infiltrates on imaging, developing mainly in women and in atopic subjects. The disorder is highly responsive to oral corticosteroid therapy, but relapses are frequent on reducing or stopping treatment. The long-term course of the disease and data regarding outcome, particularly the need for prolonged oral corticosteroid therapy and the development of severe asthma, are somewhat contradictory. A multicentric retrospective study was conducted in an attempt to describe better the initial features and, above all, the later course of CEP in a large homogeneous series of 62 stringently selected patients of whom 46 were followed for more than 1 year. The prevalence of smokers was low (6.5%) and about half of our patients (51.6%) had a previous, and often prolonged, history of asthma. The clinical and roentgenographic features were in keeping with previous studies, but we found that computed tomography could disclose ground glass opacities not detected by X-ray, and that migratory infiltrates before treatment were more frequent (25.5%) than reported previously. The bronchoalveolar lavage cellular count always showed a striking eosinophilic pattern, thus allowing distinction between CEP and cryptogenic organizing pneumonia, both syndromes sharing many common clinical and imaging features. About two-thirds of the patients (68%) showed a ventilatory defect in pulmonary function tests, with about one-half of these presenting with an obstructive pattern, sometimes without previous asthma. Along with the submucosal eosinophilic infiltration noted in 2 patients without ventilatory defect, this is strong evidence to confirm that CEP is not only an alveolointerstitial but also an airway disease. The dramatic response to oral corticosteroid therapy was observed in all treated patients. Although only 1 patient initially treated for less than 6 months did not relapse, longer oral corticosteroid therapy in no way provided protection from further relapses. We thus propose to try to wean oral corticosteroid therapy after 6 months in patients without severe asthma, because recurrences remain responsive to oral steroids. However, prolonged oral corticosteroid therapy was necessary in the majority of patients, with 68.9% of those followed for more than 1 year still on oral corticosteroid therapy at the last follow-up, either because of relapse or because of severe asthma.