Two methods for predicting regions susceptible to invasion by alien aquatic animals were developed for southern Africa (excluding Zimbabwe and Mozambique). In the "traditional" (data-poor) approach, distributions of three categories of alien "indicator" species (warm mesothermal, cold stenothermal and eurytopic) were compared to seven existing biogeographical models of distribution patterns of various animals in southern Africa. On the basis of these comparisons a synthesis model was developed which divided southern Africa into seven regions characterised by their susceptibility to invasion by alien aquatic animals with particular habitat requirements. In the "data-rich," geographic information systems (GIS) approach, the distribution of trout (Oncorhynchus mykiss and Salmo trutta) in selected "sampled regions" was related to elevation (as a surrogate of water temperature) and median annual rainfall (MAR) (as a surrogate of water availability). Using concentration analysis, optimum conditions for trout were identified. Regions within a larger "predictive area" which satisfied these conditions, were plotted as a digital map using the IDRISI package. Using this method seven models of potential trout distribution were generated for the following regions: northern Natal (two); southern Natal/Lesotho/Transkei (three), eastern Cape (two) and western Cape (two). Since two of the models were used to refine the methods, only five models were considered for the final assessment. In a modification of the GIS method, another model of potential trout distribution, based on mean monthly July minimum air temperature and MAR parameters, was developed for the region bounded by 29º - 34º S and 26 º - 32°E. This model showed marked similarities to another model, developed for the region bounded by 29 º - 32°S and 26º - 32°E, which was based on elevation and MAR parameters. The validity of the models developed was assessed by independent experts. Of the six models considered, four received favourable judgements, one was equivocal and one was judged to be poor. Based on these assessments it was concluded that the GIS method has credibility and could be used to develop a "data-rich" model of the susceptibility of southern Africa to invasion by alien aquatic animals. This method represents an alternative to the bioclimatic matching approach developed by scientists in Australia. The GIS method has a number of advantages over the "traditional" method: it is more amenable to testing, has greater flexibility, stores more information, produces images of a finer resolution, and can be easily updated. The traditional method has the advantage of being less expensive and requiring a less extensive database.