The relationship between atopy and pulmonary function in children, and how these relate directly or indirectly to airway hyperresponsiveness, is uncertain. We have examined these relationships in a sample of 13-year-old children. A questionnaire on respiratory symptoms, skin-prick tests to 11 common allergens, spirometry and an abbreviated methacholine challenge test were completed by 662 members (341 boys) of a birth cohort of New Zealand children followed longitudinally to age 13. There was a significant relationship between the presence and degree of atopy, and baseline pulmonary function. Low FEV1/VC ratios were associated with a greater likelihood of airway responsiveness, not only in subjects with diagnosed asthma, but also in the full cohort and in the sub-group of 426 children who denied asthma or current wheeze. The relationships between baseline FEV1/VC and airway responsiveness were stronger in atopic than in non-atopic children, with the strongest relationships in children sensitive to house dust mite and/or cat dander. In the presence of atopy, progressively lower levels of lung function were strongly associated with a higher prevalence of airway responsiveness (P<0.001). In non-atopic subjects, only those with the most impaired lung function (FEV1/VC < 75%) showed any substantive prevalence of airway responsiveness. The relationship between the degree of atopy and the FEV1/VC ratio, although significant in univariate analysis, became completely non-significant after accounting for airway responsiveness. In 13-year-old children, atopy, especially to house dust mite and cat dander, was correlated with pulmonary function expressed as FEV1/VC ratio. Airway responsiveness likewise correlated with impaired baseline lung function. The apparent relationship of lung function to atopy occurred primarily as a result of the relationship between atopy and airway responsiveness. Atopy and impaired lung function were additive factors predicting airway responsiveness.