The use of microsatellite markers has matured and become commonplace for plant genome analyses and is now poised for widespread practical application in sugarcane. Sequence Tagged Microsatellite Site (STMS) amplification is the most prevalent microsatellite-based approach and involves the amplification of a microsatellite by designing primers that flank and hence define the microsatellite site, revealing variation in the length of repeat motifs between individuals. Twenty-six microsatellite primer pairs received from the International Sugarcane Microsatellite Consortium (ISMC) were evaluated and the STMS protocol was optimised to ensure robust and reproducible results. The objectives of this study were to use STMS for sugarcane parentage analysis and fingerprinting. Previously, Restriction Fragment Length Polymorphism (RFLP) marker data had suggested that the parentage of a genetic mapping population, sugarcane cross AA40 (N18 x CP57/614), was incorrect. Based on the assertion that the incorrect parentage was as a result of either mislabelling at planting or at seed collection, microsatellite parentage analysis was carried out on eight potential parent pairs (13 cultivars). A total of 75 markers were scored with non-parental bands (12 on average) being observed for all of the potential parent pairs and none could be identified as the true AA40 parents. It has been suggested in other plant species that PCR artefacts could give rise to non-parental bands and to investigate this the marker data of single parent DNA reactions and pooled parent pair DNA reactions or 'synthetic offspring' were compared. The results suggested that either a certain percentage of non-parental bands, perhaps 10% (maximum value observed), should be tolerated in microsatellite parentage analysis or a marker should only be considered to be discriminating for parentage if it is absent in both the parents and the pooled parent pair amplifications. Fingerprinting of 20 cultivars using 14 microsatellite primer pairs was conducted to evaluate the potential of the STMS approach for sugarcane varietal identification. It was found that only two microsatellite primer pairs were required to discriminate between all 20 cultivars with a theoretical number of non-differentiated pairs of cultivars (XK) of only 0.03. This estimator was used to determine the approximate number of microsatellites necessary for large-scale sugarcane fingerprinting. / Thesis (M.Sc.)-University of Natal, Durban, 2002.