Globally, about one-third of final energy use and associated carbon dioxide emissions are sourced from buildings, the great majority of which are located in urban areas. Not surprisingly, managing building energy demand is a focus of city-based climate change policies while simultaneously tackling issues of fuel poverty. Assessing the potential for mitigation and evaluating the efficacy of energy policies relies on knowledge of the urban building stock, which varies geographically based on the age of building and retrofits that may have taken place. However, building data at a detailed scale are rarely available for these purposes. In this research, we present a geographic building database for Dublin city centre using a typology approach. The resulting data consists of material and energy attributes of over 25,000 buildings that have been constructed over a 200-year period. These data are used to estimate the energy ratings of households and to evaluate historic and potential retrofitting. In addition to energy studies, they provide a fundamental dataset on buildings that can be used to evaluate official sources and to support a wide range of urban research topics. The methodology used here is sufficiently general in nature that it can be expanded to other cities in Ireland and Europe