The aim of this study is to find and test a predictive model that could be suitable to estimate the outdoor NO(2) concentrations at individual level, by integrating ecological measurements recorded by local monitoring stations with individual information collected by a questionnaire. For this purpose, the data from the Italian centres of the European Community Respiratory Health Survey II (ECRHS II) has been used. Outdoor NO(2) concentrations were measured using NO(2) passive sampling tubes (PS-NO(2)), exposed outdoor for 14 days, between January 2001 and January 2003. Simultaneously, average NO(2) concentrations were collected from all the monitoring stations of the three centres (MS-NO(2)). Individual measurements carried out with passive samplers were compared with the corresponding NO(2) 2-week concentrations obtained as the average of all local (background and traffic) monitoring stations (MS-NO(2)). A multiple linear regression model was fitted to the data using the 2-week PS-NO(2) concentrations as the response variable and questionnaire information and MS-NO(2) concentrations as predictors. The model minimizing the root mean square error (RMSE), obtained from a ten-fold cross validation, was selected. The model with the best predictive ability included centre, season of the survey, MS-NO(2) concentrations, type and age of building, residential area and reported intensity of heavy-duty traffic and explained the 68.9% of the variance. The non-parametric correlation between PS-NO(2) and the concentrations estimated by the model is 0.81 (95% CI: 0.77-0.85). This study shows that over short periods (2 weeks) a good prediction of home outdoor exposure to NO(2) can be achieved by simply combining routinely collected ecological data with dwelling characteristics and self-reported intensity of heavy traffic. Further studies are needed to extend this prediction to long-term exposure.