A postal questionnaire inquiring into the presence of respiratory symptoms was sent to 3698 children aged 7 and 8 years. Those reporting either cough or wheeze were randomised, and a sample invited to attend for skin testing. A total of 192 symptomatic children, half of whom were atopic and half non-atopic, were randomly selected to enter the study. All children performed and recorded best of three peak expiratory flow measurements twice a day and completed a 10 point respiratory symptom score card each day for one year. They also recorded all treatment taken and made a note of relevant life events. Each child was seen monthly for general assessment and for measurement of nonspecific bronchial responsiveness to methacholine. Of the 192 children, 183 successfully completed the study. For six subjects 11 measurements of the provocation dose required to cause a 20% fall in forced expiratory volume in one second (PD20) were available for analysis and on the remaining 177, greater than or equal to 12 measurements. The prevalence and degree of bronchial hyper-responsiveness (PD20 less than 6.4 mumol) and its relationship to atopy was examined by comparing the percentage of members of each symptom group demonstrating bronchial hyper-responsiveness and the number of occasions on which they did so, and by comparison of minimum and median PD20 values. The range of bronchial responsiveness shown during the study period by each child was expressed as doubling doses of methacholine and compared between symptom groups. Atopy and wheeze were both independently associated with an increased prevalence and greater degree of bronchial hyper-responsiveness when compared with non-atopy and cough respectively, all differences being significant a the 0.001 level.. Thirty three per cent of subjects demonstrated a rnge of methacholine responsiveness of >4 and 13.4% of >6 doubling doses during one year.