The objective of this study is to provide an explanatory model for total energy consumption in electrically heated single-family dwellings, based on publicly available socio-economic records in Sweden. An earlier study based on 3,200 houses, divided into 93 groups of similar design, has shown that energy consumption for one house may be twice that of another house in the same area, built to an identical design. The problem is: how much of this scatter depends on occupancy behaviour? The present study is based on 78 similar houses, electrically heated, built as one group in 1969.