Wheezing in early childhood covers a wide spectrum of morbidity. Since little is known about the factors determining either the pattern or the severity of this range of symptoms, 51 children, admitted to hospital with acute wheeze in the first 2 yrs of life, were monitored prospectively between the ages of 4.5-5.5 yrs. Our hypothesis was that the predictors of severe episodes and of interval symptoms in 5 year olds would differ. Symptom diaries were kept, from which the symptom pattern and severity was assessed. The frequency and severity of acute episodes were analysed separately from day-to-day (interval) symptoms. A physiological assessment was made at 5 yrs. During the 12 month study period, 11 children were symptom-free, 15 were reported to wheeze only in response to viral infections, and 25 wheezed from multiple triggers. Bronchial responsiveness was significantly increased in those with a family history of asthma but was unrelated to any index of atopy. In a multiple logistic regression analysis, a family history of asthma and a personal history of allergy (but not results of skin-prick testing or serum immunoglobulin E (IgE) significantly predicted both attack severity and interval symptoms. An additive effect of two factors (atopy plus a family history of asthma or bronchial responsiveness) on symptom severity was suggested, without any evidence of an interaction. It is concluded that in this population of 5 year olds, with an early history of severe wheezing, familial bronchial responsiveness and atopy of operated independently to determine both interval symptoms and attack severity.