This paper describes a regional energy-environment modelling approach that explicitly takes into account the spatial distribution of production and demand activities, with their associated emissions. Indeed some of the information used at regional level has a spatial representation (e.g. district heating, electricity grid, natural gas network, traffic, public transport, etc.). We have defined a model structure that allows spatial modelling through the use of integer programming techniques in association with a linkage to a geographical information system. The energy model then becomes a part of an integrated interactive Decision Support System, allowing the decision maker to assess the consequences of the energy policy choices on air pollution in its implementation phase for the Canton of Geneva, Switzerland.