In this study, 65 allergic asthmatic children were evaluated for the presence of exercise induced bronchospasm (E.I.B.). Pulmonary function (volume at 1 second of forced expiration--FEV1; maximal mid-expiratory flow--FEF 25-75; peak expiratory flow--PEF) was assessed before and serially for 20' following treadmill exercise. Only 18 children (27.6%) presented a fall in FEV1 > or = 20%. There was a relationship between severity of asthma and incidence of E.I.B., as children with episodic, frequent and chronic asthma presented respectively 13/55 (23.4%), 2/6 (33.3%) and 3/4 (75%) of positive responses. The presence of E.I.B. cannot be predicted by clinical history, as 15/36 (41.7%) of subjects with positive clinical history had E.I.B., against 3/29 (10.3%) of subjects with negative clinical history (p = n.s.). E.I.B. in asthmatic children is less frequent as reported, correlates with severity of asthma and cannot be accurately predicted from the history.