Two populations of schoolchildren, one living in an area where the predominant allergens are house dust mites and the other in an area where the predominant allergens are pollens, were studied to investigate in more detail the associations between atopy, bronchial hyperresponsiveness (BHR) and symptoms of asthma. The prevalence of atopy (39%) was the same in both towns but the prevalence of BHR was higher in the inland 'pollen' area (19% vs 15%). Atopic children had an increased risk of having BHR and, to a lesser extent, respiratory symptoms, diagnosed asthma and hay fever. The risk of BHR was further increased in children atopic to both pollens and house dust mites, and in children with a high index of atopy (derived from the number and size of the skin reactions to four allergen groups). In addition, the relationship between atopy and BHR was quantitative in that the severity of BHR increased with the severity of atopy. We conclude that there is not a direct causal relationship between atopy and BHR but there may be a common mechanism involved in their development. It appears that, within the atopic population, the type of allergen to which the individual is sensitized, the quantity of aeroallergen present in the environment and the degree of atopy, as measured by the number and size of the skin reactions, are all factors that may interact to increase the risk of BHR.