Little is known about the relation of bronchial responsiveness (BHR) to sensitization to individual allergens, or its variation between countries. Data were obtained for BHR, specific immunoglobulin E and confounding variables from 11,215 subjects, aged 20-44 yrs at the start of the European Community Respiratory Health Survey, in 34 centres in 15 countries. The relation of BHR to sensitization to cat, house dust mite, timothy grass and Cladosporium was estimated by means of multiple regression for each centre, and combined across centres by random effects meta-analysis, controlling for baseline lung function, height, sex, season of testing, age, smoking and age/sex and age/smoking interactions. BHR was greater, on average, in those sensitized to cat (p=0.023), house dust mite (p<0.001) and timothy grass (p=0.018), but not to Cladosporium (p=0.60), and increased with degree of sensitization (p<0.001). All relations showed heterogeneity between centres, although to a lesser extent in the relation to sensitization to house dust mite. More variation in bronchial responsiveness was explained by sensitization and degree of sensitization to the individual allergens than by atopy defined as any positive test in each centre, but the relative importance of each allergen varied. The use of atopy as a single variable in relation to bronchial hyperresponsiveness may be misleading.