M.A. / The purpose of the mini-dissertation is to investigate the implementation of the language policies of financial institutions. These institutions are ABSA and NEDCOR. The research was conducted against the backdrop of the 1996 Constitution. Despite the constitutional recognition of eleven official languages, the use of official languages is only enforced in national, provincial and local government institutions. The task of promoting functional multilingualism in the private sector has fallen to PANSALB. However, English is being used overwhelmingly, especially in business. Unfortunately only 25% of speakers of indigenous languages speak English well enough to actively partake in the economy. Scant research exists regarding language preferences of clients of financial institutions. Banks claim their research shows English as the preferred banking language. They are therefore reluctant to implement indigenous languages ABSA and NEDCOR were chosen for the study and research was conducted at selected branches of ABSA, Nedbank and People’s Bank. Both quantitative and qualitative research was employed and managers, ground staff and clients were included in the study. Information was gathered over a period of eighteen months, starting from February 2003. Two hypotheses were tested: a) A discrepancy exists between the theoretical and practical implementation of the language policies of financial institutions; and b) A discrepancy exists in the language behaviour of people with different educational levels. During the empirical part of the study qualitative interviews were held with representatives of ABSA and NEDCOR who deal with the language policies of their respective banks. The sentiment is expressed in each case that the bank knows that its organic growth lies in the so-called “unbanked” sector of the market, which translates into the speakers of indigenous languages. The official/unofficial policies state that there is no discrimination against any language, but that in the case of services not being available in a specific language, English will be used. The argument used for the fact that there are few services available in indigenous languages is the lack of financial terminology and lack of legal validity of documents written in indigenous languages. Research proved both arguments invalid. Banks felt that the implementation of indigenous languages would incur great costs. Ground staff and clients at all the branches answered questionnaires and were interviewed. In this way quantitative as well as qualitative data was collected. Looking at the two hypotheses that were tested, the results show that a discrepancy exists between the theoretical and practical implementation of the language policies of ABSA and NEDCOR. Far more is done by managers and ground staff at both these institutions to accommodate speakers of indigenous languages than by policy makers. The second hypothesis, which examines a discrepancy in the language behaviour of people with differing educational levels, also proves to be valid. Respondents with a low educational level need services in their mother tongue to operate but those with a high level of education, want to see their languages being used because of a cultural identification and a pride in their languages.