This article is a commentary on a study published in the journal Int Arch Occup Environ Health 2001;74:574-8. Exposure to air pollution has been linked to respiratory effects. However, the long-term effects remain controversial. Studies examining long-term exposure are therefore still needed. To analyze the association between short-term variations in pollution indicator levels and health effects, many researchers have adopted the time-series method. However, this approach is not suitable for studying the long-term and chronic effects of such exposures. Geographical studies that allow for comparing the occurrence of effects based on different pollution levels would be more appropriate. In this study, conducted in two Central European cities where SO2 levels are relatively high, the authors use a new method based on air pollution modeling coupled with a geographic information system (GIS). This method allows for the estimation of air pollutant concentration levels on a very small scale within a specific geographic area. This study examines the relationship between outdoor SO2 levels and respiratory symptoms in children. The exposure estimation method used, combined with the high levels of SO2 exposure, increases the likelihood of identifying a positive relationship if one exists. The value of GIS for this type of epidemiological analysis is a major finding of the study. However, it would have benefited from being supported by some cartographic representations, which would have allowed for a better understanding of the relationships described. The interpretation provided regarding industrial emission sources, in which dispersion effects are not mentioned at all, remains insufficient for a definitive conclusion. (Excerpts from the article)