To investigate the bronchial response to exercise, we studied a random sample of 494 children and adolescents, aged 7-16 years, from Copenhagen. Exercise challenge consisted of steady running on a 10% sloping treadmill for 6 min in a climate chamber. Furthermore, in 464 subjects a histamine challenge test was also performed. Of the 494 subjects studied, 81 (16%) had at least 10% and 30 (6%) at least 15% reduction in FEV1 within 15 min after exercise. Twenty-nine (6%) subjects had bronchial hyperresponsiveness to both histamine and exercise, 48 (10%) subjects had bronchial hyperresponsiveness to exercise, but histamine responsiveness within the normal range, whereas 340 (73%) subjects had neither bronchial hyperresponsiveness to exercise nor inhaled histamine. With regard to the presence of asthma defined as substantial exercise induced bronchoconstriction (delta-FEV1 greater than or equal to 10%), exercise testing may not be appropriate for identifying clinical asthma in a random sample, because the highest predictive value of a positive test was 25%. On the other hand, a history of clinical asthma was frequently associated with increased bronchial responsiveness to exercise (77%). In conclusion, 16% of a random sample of children and adolescents had abnormal bronchial responsiveness to exercise (delta FEV1 greater than or equal to 10%), 6% of the subjects had a delta FEV1 greater than or equal to 15%. Furthermore, because of a low predictive value of a positive test, the exercise challenge test has only a supplementary role in the detection of clinical asthma in population samples.