The residential sector plays an important role in the energy system of developing countries. In this paper we introduce a bottom up simulation model for household energy use. The model describes energy demand for several end-use functions based on a set of physical drivers, such as floor space and heating degree days. The model also recognizes different population groups: i.e. urban and rural households, each distinguishing five income quintiles. The model is applied to analyze possible future developments of residential energy use in five developing world regions: India, China, South East Asia, South Africa and Brazil. We find that in each of these regions cooking is currently the main end-use function, but that other functions, such as space heating, cooling and appliances become more important. At the same time, energy consumption slowly shifts towards modern fuels. The model also shows that climate policy can reduce residential energy emissions, but could also slow down the energy transition away from traditional fuels in low income classes.