Airway responsiveness is an objectively measurable clinical trait related to the presence of asthma. Although risk factors for this trait have been evaluated cross-sectionally, little is known about its longitudinal predictors. A population cohort of 539 children, aged 8 yrs at the start of follow-up, underwent 3-7 bronchial challenge tests spaced at 3-9 month intervals. Airway responsiveness was assessed by a single-step distilled water challenge. To investigate responsiveness as a continuous trait, the ratio of the postchallenge to prechallenge forced expiratory volume in one second was calculated. A child's repeated ratios were then regressed on the time of follow-up. The resulting child-specific regression coefficient was the outcome variable to assess longitudinal predictors of airway responsiveness (AR). Results were based on 2,267 repeated challenge tests, and indicated an overall decrease in responsiveness. Children with a diagnosis of asthma (mean+/-SD, longitudinal change in AR + yr(-1): -0.060+/-0.149), those with a positive skin-prick test (-0.018+/-0.106) and those with reported exposure to maternal smoking (-0.004+/-0.083) demonstrated increased airway responsiveness over time. All of these changes in airway responsiveness were found to be significant in multiple linear regression analysis (p<0.05). Stratified analysis further indicated that the effect of maternal smoking was observed primarily in nonatopic children and was greatest at an exposure level of > or =10 cigarettes x day(-1). In conclusion, atopy and exposure to maternal smoking can predict longitudinal change in childhood airway responsiveness.