A one year study on a random sample of 1225 Swiss children aged 0-5 years was conducted in four different areas (two urban, one suburban and one rural) of Switzerland to investigate the relationship between air pollution and respiratory symptoms. For each child daily symptoms over a six week period were recorded by their parents in the form of a standard diary and air pollution was assessed by personal NO2-samplers. 20% of the diaries were validated by comparison with the attending pediatrician's case-notes and showed good agreement (87%). The frequency of respiratory symptoms per child per day was found to increase with increasing levels of NO2 measured outdoors, but not with NO2 concentration indoors (when other indoor sources for NO2 where present). Possible other factors were accounted for by multiple regression analysis and the variables "season" and "child's susceptibility to colds" also showed a significant association with respiratory symptoms. But the relationship between NO2 outdoors and respiratory symptoms per child per day remained statistically significant. The multiple regression model explains 7% of the total variability. The result indicates that air pollution is a contributory factor in the development of respiratory symptoms in children.