To develop effective air quality management strategies at regional and urba
n levels requires the preparation of several actions belonging to different
scientific areas and analysis of their complex interactions. A well struct
ured management system must be based on efficient tools that integrate dive
rse sets of information - emissions, traffic, meteorological variables and
air quality data. Ideally, such a system should include appropriate modelli
ng tools that could be used to perform spatial interpolation and air qualit
y forecasts. This paper will describe the modelling system developed for pr
edicting the emissions (emission pre-processor), dispersion and chemical tr
ansformation of pollutants in urban areas and its application for two sites
in Portugal (Lisbon and Oporto). For predicting the concentration fields,
all categories of sources are taken into account, namely stationary point a
nd area sources and traffic. This allows the construction of an urban air q
uality management system for decision support.