Access to clean, affordable and appropriate energy is an important enabler of development. Energy allows households to meet their most basic subsistence needs; it is a central feature of all the millennium development goals (MDGs) and, while a lack of access to energy may not be a cause of poverty, addressing the energy needs of the impoverished lets them access services which in turn address the causes of poverty. While much is known about the factors affecting the decisions made when choosing between fuel types within a household, few quantitative studies have been carried out in South Africa to determine the extent to which these factors affect energy choice decisions. It is assumed that the factors traditionally included in economic demand such as price and income of the household affect choice; tastes and preferences as well as external factors such as distance to fuel suppliers are expected to influence preferences. This study follows two typical low-income rural sites in South Africa, Antioch and Garagapola, where the Electricity Basic Services Support Tariff (EBSST) was piloted in 2002. The EBSST is set at 50 kWh/ month per household for low domestic consumers; this is worth approximately R20 1 (7US$3). This subsidy is a lifeline tariff, where households receive the set amount of units per month, free of charge irrespective of whether more units are purchased. These data (collected in 2001 and 2002), recently collated with detailed electricity consumption data, allow us to determine the drivers of electricity consumption within these households. The sample analysed is taken from the initial phase of the study, when no FBE had been introduced to the households. This enabled the study presented here to make use of the well-populated datasets to assess what affects the electricity use decision in these households. This paper attempts to assess which factors affected the decision-making process for electricity consumption within these households. A brief history of the electricity industry and the electrification is provided and the theoretical background for the electricity consumption model is provided. It was found that income, woodfuel usage, iron ownership and credit obtained were significant in determining consumption levels within these households. Price and cross-price elasticities were difficult to assess due to lack of data within the sample. The results have many possible implications for policy, including the effect that easily obtained credit has for low-income households.