The physiopathology of chronic cough remains obscure. We evaluated the possibility that chronic cough in nonasthmatic subjects is associated with airway inflammation, and if this is so, what the relationship between this inflammation and the possible etiology of cough might be, as well as its response to inhaled steroids. Nineteen nonsmoking, nonasthmatic subjects referred for a persistent cough (mean: 3.8 yr) were evaluated and compared with 10 normal subjects. The evaluation included a respiratory questionnaire, a physical examination, allergy skin-prick tests, chest and sinus radiographs, esophageal pH monitoring, measurements of expiratory flows, methacholine and citric acid challenges, and flexible bronchoscopy for bronchoalveolar lavage (BAL) and bronchial biopsies. Fourteen subjects further accepted participation in a randomized, double-blind crossover trial of inhaled beclomethasone (500 micrograms four times daily) and a placebo for 1 mo each. Four groups of subjects were identified according to the presence of postnasal discharge (n = 4), gastroesophageal reflux (n = 6), both conditions (n = 5), or neither (n = 4). Subjects with chronic cough had an increased number of inflammatory cells in their bronchoalveolar lavage fluid (BALF), but there was no significant difference between the four subgroups of coughers. As compared with control subjects, the bronchial biopsies of subjects with chronic cough showed increased epithelial desquamation (p = 0.004) and inflammatory cells (p = 0.005), particularly mononuclear cells (p < 0.01), in addition to submucosal fibrosis, squamous-cell metaplasia, and loss of cilia. These findings were not significantly different between the different etiologic groups. In subjects with chronic cough, basement-membrane thickness was normal and not different from that of control subjects.(ABSTRACT TRUNCATED AT 250 WORDS)