Elevated concentrations of nitrogen dioxide (NO2) are produced in the home by the use of unvented gas appliances. In studies on potential health effects of indoor exposure to NO2, exposure has mostly been estimated from the presence or absence of sources like gas cookers in the home. This leads to misclassification of exposure, as NO2 concentrations in the home depend also on source use, ventilation habits, time budgets, etc. The availability of cheap, passive monitoring devices has made it possible to measure indoor concentrations of NO2 directly in health effects studies, albeit with averaging times of one to several days. So far, it has not been evaluated whether this increases the sensitivity of a study to detect health effects of NO2. In this paper, a comparison is made between NO2 sources and weekly average indoor NO2 measurements, as predictors of pulmonary function in a study among children aged 6-12 years. The relationship between exposure and lung functions was found to be generally non-significant in this study. The results further suggested that in this study, measuring indoor NO2 concentrations with passive monitors offered no advantage over the simple use of source presence as exposure variable.