The effect of passive smoking on the degree of nonspecific bronchial responsiveness (BR) in children was evaluated. In a cross-sectional survey of primary school children (7 to 11 yr) in Latium (Italy), a total of 1,215 methacholine challenge tests was performed. The children were divided into nonresponders (PC20 FEV1 > 64 mg/ml) and responders (PC20 FEV1 < or = 64 mg/ml), the latter including a subgroup of strong responders (PC20 FEV1 < or = 4.0 mg/ml). The role of passive smoking exposure was analyzed through logistic regression models. Higher odds ratios were found among girls (for strong responders: maternal smoking, OR 2.92; paternal smoking, OR 2.59); moreover, among girls there was a dose-response relationship with the number of cigarettes smoked by the mother. An effect modification was also detected for father's education and household crowding: maternal and paternal smoking were strong predictors of bronchial hyperresponsiveness in families in which the head was less educated and in overcrowded houses. The results were confirmed using a continuous index of BR (slope) and after adjustment for baseline airway caliber. We suggest that the effect of parental smoking on children's bronchial responsiveness is detectable when the conditions for a higher exposure level at home are met; females seem to be more susceptible.