Florey C du V [Department of Community Medicine, St Thomas's Hospital Medical School, London SE1 7EH], Melia R J W, Chinn S, Goldstein B D, Brooks A G F, John H H, Craighead I B and Webster X. The relation between respiratory illness in primary schoolchildren and the use of gas for cooking. III Nitrogen dioxide, respiratory illness and lung function. International Journal of Epidemiology 1979, 8: 347–353. We examined the relation between lung function and respiratory illness in a population of 808 primary school children aged 6–7 years and the levels of nitrogen dioxide (NO2) in the kitchens and bedrooms in their homes. Complete data were collected on about 66% of the population. The children lived in a defined 4 square km area in Middlesbrough, Cleveland, UK. One week average outdoor levels of NO2 varied little over the area (14–24 ppb). The prevalence of respiratory illness was higher in children from gas than electric cooking homes (p ≏ 0.1). Although prevalence was not related to kitchen NO2 levels (range 5–317 ppb) it increased with increasing levels of NO2 in the children’s bedrooms in gas cooking homes (range 4–169 ppb, p≏0.1). Symptoms in siblings and parents were not related to kitchen NO2 levels. Lung function was not related to NO2 at which an association with illness was observed and the inconsistency between our results in the UK and those from several studies in the US, it is possible that the NO2 levels were a proxy for some other factor more directly related to respiratory disease such as temperature or humidity.