The outcome after 25 yr was studied for three groups of children classified in a random community survey in 1964 as having asthma (121 subjects), wheeze in the presence of infection (167 subjects), or no respiratory symptoms (167 comparison subjects). Approximately 80% of the subjects in each group, now aged 34 to 40 yr, were successfully traced. Current symptoms and smoking habit were recorded by questionnaire, and ventilatory function, peak flow variability, and bronchial reactivity to inhaled methacholine were measured. Subjects who had asthma in childhood were more likely to wheeze (odds ratio [OR] 14.4) or produce phlegm (OR 3.3) than comparison subjects. They also had significantly lower FEV1 values and greater bronchial reactivity than comparison subjects. Adult FEV1 correlated with childhood FEV1 (both expressed as % of predicted) (r = 0.44, p < 0.01). The prognosis for those children who were classed as having wheeze in the presence of infection in 1964 was better than for those who had asthma. Although they also were more likely to report wheeze (OR 3.8) or phlegm (OR 4.4) than comparison subjects, the wheezy symptoms were unlikely to interfere with activities and the ventilatory function and bronchial reactivity to methacholine did not differ from those of comparison subjects. Smokers were more likely to report wheeze (OR 2.0), cough (OR 7.2), and phlegm (OR 3.1) than never-smokers, and current smokers with current wheezy symptoms had significantly reduced FEV1 values, although smoking was not associated with increased methacholine reactivity.(ABSTRACT TRUNCATED AT 250 WORDS)