Radon causes lung cancer when inhaled for prolonged periods of time. A range of factors influence residential radon concentration and this study therefore sought to ascertain which dwelling-related factors exert an influence on radon levels. A cross-sectional study was conducted from 2001 to 2003 which analyzed 983 homes of as many subjects randomly selected from the 1991 census. Sampling was carried out by district and stratified by population density to ensure that more detectors were placed in the most heavily populated areas. Radon concentration and different dwelling characteristics were measured in each of the homes selected. Bivariate and multivariate analyses were performed to ascertain which factors influenced radon concentration. The geometric mean of radon concentration was 69.5 Bq/m3, and 21.3% of homes had concentrations above 148 Bq/m3. Factors shown to influence radon concentration in the bivariate analysis were: age of dwelling; interior building material; exterior building material; and storey on which the detector was placed. Explanatory variables in the multivariate analysis were: age of dwelling; number of storeys; distance off floor; and interior building material. The model was significant, but the variability explained was around 10%. These results highlight the fact that the study area is an area of high radon emission and that factors other than those directly related with the characteristics of the dwelling also influence radon concentration.