This study was conducted to see whether children living in socially deprived areas were more likely than other children to be admitted to hospital for asthma, and, if so, whether their excess risk was attributable to a higher prevalence of asthma or poorer treatment. Hospital admission rates for asthma were obtained for Cardiff electoral wards and compared with the Townsend indices of deprivation. A survey of respiratory symptoms was conducted in schoolchildren; prevalence of symptoms was compared with Townsend index and asthma admission rate for the schools' catchment areas. Asthma admissions were strongly correlated with Townsend indices at all ages. The prevalence of reported asthma and various degrees of wheeze in the schools was not significantly correlated with Townsend index or hospital admission rate in the corresponding areas. The presence of a smoker in the house was strongly associated with Townsend index and admission rate; children whose houses contained a smoker were more likely than others to have wheezed in the past year and to have disturbed nights due to wheezing. There was a non-significant negative association between Townsend index and regular use of inhaled steroids. The relationship between hospital admission for asthma and social deprivation is not explained by variations in prevalence, but it may be attributable to the aggravation of symptoms by active or passive smoking, and perhaps also to differences in management.