This study evaluated the prevalence of increased bronchial responsiveness (BR) in children living in two areas with different air pollution levels. A total of 1,215 methacholine challenge tests were performed among a random sample of primary schoolchildren in an industrial town polluted by oil-fired thermoelectric power plants and in a rural area chosen as a control. The two groups showed similar lung function data (forced vital capacity [FVC] and forced expiratory volume in 1 s [FEV1.0]) and prevalence of positive prick tests to common aeroallergens. All children responding with a 20% drop in FEV1.0 to a methacholine concentration < or = 64 mg/ml, "all responders" (PC20FEV1.0 < or = 64 mg/ml), and the subgroup of "strong responders" (PC20FEV1.0 < or = 4 mg/ml) were compared separately with "nonresponders" (PC20FEV1.0 > 64 mg/ml), calculating odds ratios (ORs). There were more "all responders" (57.2% versus 41.4%) and "strong responders" (20.0% versus 14.8%) among subjects living in the industrial area; the excess remained when several potential confounders were taken into account ("all responders": OR = 2.0, 95% confidence interval [95% CI] = 1.5-2.6; "strong responders": OR = 1.9, 95% CI = 1.3-2.8), and it was even more pronounced among girls. When the effect of clinical predictors of increased airways responsiveness (history of asthma, allergic rhinitis, baseline airways caliber, skin reactivity, recent respiratory infection) was considered through multiple logistic regression, the odds ratios associated with living in the industrial area were still significantly increased in girls. The cross-sectional approach and the lack of individual exposure data limit the interpretation of the findings.(ABSTRACT TRUNCATED AT 250 WORDS)