BACKGROUND: The process of industrialization of the USSR has left a legacy of widespread and often poorly controlled pollution which is widely believed to have adverse implications for health, in particular for respiratory disease among children. OBJECTIVES: To assess the relationship between area of residence and respiratory function in junior schoolchildren in different districts of Moscow. METHODS: A survey was conducted of 539 children aged 6-12 years who attend school and live in one of three districts of Moscow with varying ambient pollution levels. Spirometry [forced vital capacity (FVC) and forced expiratory volume in 1 s (FEV1)] was assessed at school by trained school health staff. Parents of the children completed a questionnaire asking about respiratory function and factors potentially associated with it, as well as about social and other factors that could influence respiratory development and the health status of their children. RESULTS: There was appreciable difference in the characteristics of the children from the three districts. Children from the lower pollutant districts were generally younger, had higher parental income, and were less frequently exposed to cigarette smoke at home. They were also less likely to report heavy lorry traffic in the streets outside their homes. After adjustment for age, gender and height the FVC was 7.6 per cent (3.6-11.5 per cent) lower in children from the medium pollution district and 9.9 per cent [95 per cent confidence interval (CI) 5.6-14.0 per cent] lower in children from the high pollution district compared with those in the least polluted district (p < 0.001 for trend). These differences were little affected by further adjustment for household income or exposure to household smoking. In contrast, FEV1 showed comparatively little variation across districts. The odds of a forced expiratory ratio (FER) <75 per cent were substantially lower in the high pollution compared with the low pollution district (odds ratio 0.10, 95per cent CI 0.03-0.32 after adjustment for age, gender and height), and there was clear evidence of a trend across pollution categories (p < 0.001). The frequency of reported allergy was also lower in the high pollution district. FVC increased, and the probability of a low FER decreased, with household income. CONCLUSION: Children from areas of high environmental pollution had lower lung capacity but also smaller risk of a low FER compared with those from cleaner areas. The extent to which these differences can be attributed to environmental pollution is unclear without more detailed study. However, socio-economic deprivation, which was associated with pollution, appears to be an important determinant of respiratory function although it was associated with a lower risk of an obstructive pattern of lung function tests.