Models used for forecasting of future energy demand are often econometrically based and do not attempt to look in detail at end uses of energy. It has become clear in recent years that 'this approach to energy forecasting can produce more realistic estimates of future demand in the building sector if it is used in conjunction with models which-cover important physical and social parameters and the contraints which these impose (Grubb 1990; Grubb 1991). Such complementary models provide insights which would otherwise be missed. The BREHOMES model meets the requirements of such a complementary model for one particular part of the building sector - it is a physically based model of energy use in the United Kingdom housing stock which also takes account of social aspects. This paper explains how BREHOMES works, describes the data which it uses and presents examples of how the model is being used in practice.