Information on species diversity is urgently needed but often unavailable. Conservation biologists have therefore used the 'Higher-Taxon Approach' to predict the distribution of a-diversity of species from the diversities of higher taxa. We expanded this approach. We studied the Kenyan woody plant flora on a 56 x 56 km grid basis and found that the distribution of species a-diversity could be predicted precisely even from ordinal diversity. In contrast, the distribution of species beta-diversity was difficult to predict even from the genus level, although within biogeographic zones the prediction improved. The spatial scale of species alpha-diversity could be predicted precisely only from genus level, and was hardly predictable from ordinal level. The spatial scale of species beta-diversity was roughly predictable from genus to ordinal levels. Overall, the Higher-Taxon Approach could be expanded to a rough prediction of the spatial scale of species diversities. But, the approach only poorly predicted the spatial distribution of species beta-diversity. (