This paper focuses the novel GIS-related aspects of a prototype planning support system for urban planners and energy advisers. The system design combines commercial database and GIS packages to provide a flexible means of predicting the solar energy potential and energy consumption of dwellings. By incorporating customised tools designed to derive useful data directly from digital maps and aerial photographs, it addresses the significant problem of data collection for urban-scale energy modelling. In conjunction with data generated by a new dynamic default data system, the map-derived data provide the necessary input for an embedded domestic energy model and a solar energy calculation engine. In this way, robust estimates can be obtained, at almost arbitrary scales of resolution, of the potential for solar energy to replace delivered domestic energy and to mitigate emissions of carbon dioxide.