Focused procedures which streamline and optimise plant prioritisation and selection in bioprospecting have the potential to save both time and resources. A variety of semiquantitative techniques were assessed for their ability to prioritise ethnomedicinal taxa in the Flora of Southern Africa (FSA) region. These techniques were subsequently expanded upon for application in plant selection for the Novel Drug Development Platform bioprospecting programme. Least squares regression analyses were used to test the hypothesis that ethnomedicinal plant use in southern Africa is strictly random, i.e. no order or family contains significantly more medicinal plants, than any other order or family. This hypothesis was falsified revealing several 'hot' plant orders. The distribution of southern African ethnomedicinal taxa was investigated, and revealed low ethnomedicinal plant usage in the Western Cape and Northern Cape. The historical settlement of Bantu tribes in the eastern regions of southern Africa was one explanation for this discrepancy. Growth forms of ethnomedicinal taxa in 'hot' orders (identified in the regression analysis) were analysed. The results indicated no clear preferences across orders, but rather a preference for particular growth forms in certain orders. With respect to distribution, endemism and Red Data List status of ethnomedicinal taxa, the Western Cape had the greatest proportion of endemics and Namibia had the highest proportion of Red Data Listed ethnomedicinal taxa. With respect to chemotaxonomy, the Asteraceae contained the highest proportion of terpenoids, the Rubiaceae the highest proportion of alkaloids and the Fabaceae the highest proportion of flavonoids. The predictive value of regression analyses was tested against an existing analysis of anti-malarials and the subsequent in vitro bioassays on Plasmodium falciparum. In particular, the ability of these analyses to identify plants with anti plasmodial IC50 values of [less than or equal to] 10 [micro]g/ml was assessed. Most species in 'hot' genera showed comparatively good antiplasmodial activities (IC50 [less than or equal to] 10 [micro]g/ml). Plant candidates were prioritised for screening anti-tuberculosis, anti-diabetes and immune-modulatory compounds, using a weighting system based on; their ethnomedicinal application, chemotaxonomic potential, frequency in ethnomedicinal trade, association with the relative disease, toxicity, Red Data status, indigenous or endemic status, and family selection in ethnomedicine (identified through regression analyses). Other taxa were short-listed due to their presence in biodiversity hotspots where few ethnomedicinal plant use records are documented, and still others were incorporated due to their taxonomic association with efficacious exotic allies. Statistical analyses of the weighting processes employed were not possible in the absence of screening results which are due only in December 2006. The legislation governing bioprospecting in South Africa is discussed and several recommendations are presented to minimise negative impacts on the industry. / Thesis (M.Sc.)-University of KwaZulu-Natal, Pietermaritzburg, 2005.