Meaningful studies of asthma genetics require careful definition of airway hyperresponsiveness (AHR). In children, several studies have emphasized the need for correction of bronchial challenge data for baseline parameters, such as age, gender, lung function and atopic status, when undertaking airway responsiveness measurements. However, few studies have suggested how this should be performed in practice. This study describes a method for the correction of dose-response slopes (DRS) and PC20 values for baseline parameters in children, and illustrates the effect of such corrections on the association of AHR with the glutathione S-transferase GSTP1 Ile105Val polymorphism in children. Skin prick and methacholine challenge testing, measurement of total serum IgE concentration and GSTP1 genotyping were performed in 145 unrelated British children aged 7-18 years. Correction of bronchial challenge results, expressed as both DRS and PC20 values, for age, gender, baseline lung function and atopic status was performed using linear regression and discriminant analysis, respectively. Adjusting bronchial challenge results for the age and size of the child altered AHR status, defined as a PC20 methacholine <8 mg/ml, in 37% of children. Correction for baseline parameters also resulted in a significant reduction in mean DRS (original uncorrected DRS 83.6, corrected DRSc 27.4). This had a marked effect on the results of the association study, unmasking a previously unidentified association between the GSTP1 genotype and AHR in children. Age and size adjustment of bronchial challenge data has a significant effect on AHR status and may influence the results of genetic association studies in children.