The aim of this study was to compare the incidence of community-acquired Legionnaires' Disease in Nottingham with England and Wales and to explore reasons for any difference observed. Based on data from the National Surveillance Scheme for Legionnaires' Disease (1980–1999), the rate of infection in England and Wales was 1·3 per million/year compared with 6·6 per million/year in Nottingham. Domestic water samples were obtained from 41 (95%) of 43 Nottingham cases between 1997 and 2000. In 16 (39%) cases, Legionella sp. were cultured in significant quantities. Proximity to a cooling tower was examined using a 1[ratio ]4 case-controlled analysis. No significant difference in the mean distance between place of residence to the nearest cooling tower was noted (cases 2·7 km vs. controls 2·3 km; P=0·5). These data suggest that Nottingham does have a higher rate of legionella infection compared to national figures and that home water systems are a source.