Candida species were discovered more than a century ago as a causative organism of oral thrush. In HIV patients, the presence of oral candidiasis has been shown to be the earliest opportunistic infection. Candidiasis lesions associated with HIV infections are primarily a reflection of the specific change of the host's immune response caused by the virus. Studies of AIDS all over the world show that 58-81% of all patients contract a fungal infection at some time during the primordial stage or after developing AIDS and 10-20% have died as a direct consequence of fungal infections. Twenty four South African medicinal plants were screened using a modification of the NCCSL broth microdilution antifungal test against Candida albicans standard strain ATCC 10231 and two clinical isolates from a 5-month- old baby and an adult. This assay was performed in order to find a traditional remedy to treat oral candidiasis. Of all the screened plants Allium sativum L., Glycyrrhiza glabra L., Polygala myrtifolia L. and Tulbaghia violacea L. aqueous extracts were found to have the best activity. Allium sativum and Tulbaghia violacea aqueous bulb extracts had MIC values of 0.56 mgml-1 and 3.25 mgml-1 respectively, whilst Polygala myrtifolia leaf extracts and Glycyrrhiza glabra rhizome extracts had MIC values of 1.56 mgml-1 and 3.25 mgml-1 respectively when tested against the isolate from a 5-month-old baby, which was the most susceptible of the isolates used. All the extracts had higher MIC values against the standard strain (ATTC 10231), which was the least susceptible to the extracts used. Stability testing was performed on fresh aqueous extracts of A. sativum, G. glabra, T. violacea and P. myrtifolia stored at 4°C, 23°C and 33°C over a period of one week, to determine the stability of the extracts in solution. All A. sativum extracts maintained stability for three days in solution, whilst T. violacea extracts remained stable for only two days in solution. TLC fingerprinting of A. sativum and T. violacea extracts indicated the presence of the known antibacterial and antifungal compound allicin. The activity of allicin and other active compounds was observed by using the bioautographic assay, which was performed on these extracts. P. myrtifolia and G. glabra extracts lost stability 24 hours after preparation at all tested temperatures. However, it was clear with the four plant extracts tested that storage of solutions at higher temperatures reduced their activity and stability. The unpleasant taste and smell of A. sativum and G. glabra could however not be masked, since the intake of these two extracts would result in HIV patients being recognised. These two plants where therefore not considered for further investigation. G. glabra and P. myrtifolia are both saponin containing plants. These could be the active constituents responsible for the anticandidal action. G. glabra is known for its biological activity as an antibacterial agent, whilst other Polygala species have been reported to possess antifungal saponins. Although P. myrtifolia and G. glabra are not stable for more than 24 hours, they do not have an unpleasant smell or taste. These plants are therefore further investigated for use as oral mouthwash in clinics and homes. / Thesis (M.Sc.)-University of Natal, Pietermaritzburg, 2003.