Fire is generally used as a management tool for different vegetation types such as savannas and grasslands in southern Africa. In the False Thornveld of the Eastern Cape, fire is commonly used to control bush encroachment and to increase grass production, as grasses are important source of forage for domestic livestock. At the University of Fort Hare farm in the Eastern Cape, a trial was set up in 1980 to investigate the effect of burning frequency on vegetation. There are six treatments replicated twice in a completely randomized design on a 100 m x 50 m plots. The treatments comprise no burn, annual, biennial, triennial, quadrennial and sexennial burns. From this trial a study was conducted to investigate long term effect of burning frequency on species and soil seed bank diversity. Two 100 m line transects located 25 m apart were laid within each plot, and the herbaceous and woody species were identified and recorded along the line transects. Relative abundances (%) for each species were calculated for each treatment. Soil samples were collected at an interval of 13.3 m along the line transects. The samples were placed in paper bags and kept for use in a germination experiment. The seedling emergence germination method was used in the glasshouse to determine emerging seedlings, pots were filled with Hygromix growth medium and the soils from the fire trials 12 X 6 = 72 were spread on top. Soil from the control plots were also pre-treated with heat, smoke and the combination of heat and smoke. The experiment started in January 2010 and was terminated in April 2010. The Shannon-Weaver Diversity Index was used to determine species diversity for standing vegetation and germinated seedlings. Data were tested for normality and species abundances were transformed. Analysis of Variance (ANOVA) was applied to test treatment effects on geminated seedlings, species abundance and diversity at α = 0.05 significance level. Significant differences between treatment means were determined by post v hoc tests using Fischer‟s Least Significant Difference test at α = 0.05. The Pearson Moment Correlation test was used to test the relationship between vegetation and soil seed banks. Treatments had significant effects on herbaceous, woody species abundance and diversity (P < 0.05), but there was no significant treatment effect on soil seed bank diversity and on pre-germination treatments (P > 0.05). The annual, biennial and triennial burns were dominated by Themeda triandra while the quadrennial, sexennial and the control treatments were dominated by Sporobolus fimbriatus and Melica decumbens. A. karroo dominated the woody component across all treatments, but it did not change in abundance. S. fimbriatus was the most abundant in the soil seed bank across all the treatments and also in all pre-germination treatments. Species diversity was high in standing vegetation in the quadrennial, sexennial and the control treatments. There was no significant correlation between the standing vegetation and soil seed bank diversities (P > 0.05). Based on these findings it is apparent that fire can change vegetation in an area to be dominated by fire tolerant or fire intolerant species. Fire frequency, heat and smoke affects soil seed banks to a lesser extent in the False Thornveld. For the management of the False Thornveld of the Eastern Cape, less frequent burning is recommended as it increases aboveground species diversity.