This paper analyses the characteristics of low-income savers in a working class residential area of Cape Town. It uses the Khayelitsha and Mitchell’s Plain Survey that was conducted in 2000. The survey was done at both a household and individual level for all adults over 18 years old. These household and individual datasets were merged to form the dataset used in this study. There were 4984 respondents of which 2644 were adults .The KMPS data set is a good foundation for analysing the characteristics of savers in a low income area characterised by high unemployment and poverty. It is sometimes argued that the South African poor do not save as every increment of their income is needed for their day-to day survival (Matin et al, 2002, 276). Others have shown that the poor do indeed save, albeit often in small amounts (Porteous 2003, Verhoef 2002). This study comes to similar findings, but goes further than the existing literature by analysing different forms of savings. Recently there has been some debate surrounding the banking system in South\nAfrica. The biggest criticism has been that the South African banking system does not provide basic saving services to the poor. 17.6 million adults in South Africa do not have a bank account (Porteous.2003. 2). Within the decile of the population with the lowest standard of living only 8% had a bank account, whereas within the 10% of the population with the highest standard of living, 91% had access to a bank account (ibid). This study explores whether these national trends are reflected in the savings patterns of the respondents in the KMPS survey. As will be seen later, although the majority of the respondents do not use the formal banking sector, the poor do save through informal means such as stokvels, gooi-goois or burial societies.