Two cross sectional surveys, 24 years apart, using the same respiratory questionnaire, were carried out to examine changes in prevalence rates of cough, phlegm, and wheeze and to relate changes in wheeze to objective peak expiratory flow rates (PEFRs). The surveys were done in towns in southern and northern England and South Wales in schoolchildren aged 6.0-7.5 years; n = 1655 in 1966 and n = 2323 in 1990. Parents reported on winter cough and winter phlegm (early morning or day/night) and wheeze; PEFRs were also measured. The proportion of children reported as wheezing on most days or nights increased from 3.9% to 6.1% (95% confidence interval (CI) for increase -0.2 to 4.6), with a smaller increase in the prevalence of those who had ever wheezed. The proportion of children with day or night time cough increased from 21.1% to 33.3% (95% CI for increase 3.8 to 20.6) and the proportion with day or night time phlegm increased from 5.8% to 10.0% (95% CI for increase 0.4 to 8.0). Smaller increases in the prevalence of persistent cough (from 9.0% to 12.4%) and persistent phlegm (from 2.4% to 3.5%) were also observed, while morning cough and morning phlegm showed little change. The increases in cough and phlegm were apparent in subjects with and without a history of wheeze. Both absolute and proportional changes in symptom prevalence were generally greater in the north than in the south. Similar social class trends were seen in each survey. The mean difference in PEFR between subjects with and without wheeze was smaller in 1990 than in 1966, but this result could be influenced by a greater proportion of subjects receiving antiasthmatic treatment in the 1990 survey. These apparent increases in the prevalence of persistent wheeze, day and night time cough and phlegm, occurring over a period during which outdoor air pollution levels have decreased substantially, deserve further investigation.