The distribution of the Unionoida is almost cosmopolitan and reaches its greatest diversity in North America with 860 currently recognized valid species. Two genera of the family Unionidae, Unio and Coelatura, comprising four species, occur in South Africa. This article focuses on the distribution and habitats of Unio caffer Krauss, 1848 based on the records in the database of the National Freshwater Snail Collection (NFSC) of South Africa. This bivalve is considered to be endemic to South Africa and although it was sporadically reported from elsewhere in South Africa, the Western Cape is the only province from which no samples are on record in the database of the NFSC. The majority of the 58 samples on record was recovered from rivers (32.8%) and dams (20.7%) and from water conditions described as perennial, clear and fresh and 22 of the samples were collected in water bodies with a predominantly sandy substratum. A temperature index calculated for this species ranked it in fi fth position of the 12 bivalve species represented in the database on account of its association with low climatic temperatures. An integrated decision-tree analysis indicated that temperature, substratum and water bodies per se were the most important factors of those investigated that played a signifi cant role in establishing the geographical distribution of this species in South Africa. Comprehensive surveys for freshwater molluscs conducted by state and local health authorities were discontinued during the eighties of the previous century and the majority of sampling sites have not been revisited since. Therefore hardly any recent data pertaining to the conservation status and species diversity of the mollusc fauna of South Africa are available. However, during relatively recent surveys conducted by the authors at three previously positive sites for U. caffer no specimens of this species could be recovered and it is also reported in literature that its range in the south-western Cape has decreased in recent years. With regard to its conservation status, the above findings seem to suggest that U. caffer should at least be considered as vulnerable – if not endangered – as reported for some related species elsewhere in the world. Although speculative, several reasons are suggested to explain the global phenomenon of decline in freshwater bivalves. These include, amongst others, construction of impoundments, introduction of alien species, wetland drainage and canalization and pollution. However, the unique lifecycle of the Unionoida could also play an important role in this respect due to the fact that their larval stages are obligatory parasites on fish. These bivalves are therefore dependent on fish for their survival and dispersal and without their host fish populations will disappear. To sustain a viable population a water body should therefore be suitable not only for the bivalves themselves but also for their host fi sh. As mentioned earlier, the majority of samples of U. caffer were recovered from dams and rivers, water body types both under pressure of over exploitation and pollution. It is therefore recommended that thorough surveys should be planned and conducted in specific areas which could be selected with the documented geographical distribution in the database of the NFSC as guideline. A comparison of the results of such surveys with the data in the database of the NFSC could make a considerable contribution towards assessing the current conservation status and diversity of the freshwater molluscs of South Africa.