The Greater Cape Floristic Region (GCFR) of South Africa has one of the most biodiverse floras in the world. Although ecological speciation is thought to be a primary mechanism behind diversification in the GCFR, it has recently been hypothesised that non-ecological speciation may have an influential role in driving diversification in the montane “sky islands” of the GCFR, potentially resulting in cryptic species. This work seeks to test the relative importance of ecological versus non-ecological processes in powering speciation at different elevations in the GCFR, and to assess the existence of multiple, range-restricted cryptic species at high elevations. The Cape grass genus Ehrharta is an ideal system in which to investigate these processes, as it contains low-, mid-, and high-elevation lineages, with a previously documented adaptive radiation in the succulent karoo. Population-level phylogenetic analyses using targeted enrichment sequencing data show that E. rupestris and E. setacea, as currently defined, are polyphyletic, and reveal multiple distinct monophyletic lineages within the Ramosa, Rehmannii and Setacea clades of genus Ehrharta. Analysis of genotyping-by-sequencing and morphological data, together with evidence of sympatry, confirm these lineages to be distinct entities, resulting in 13 to 16 putative new species, of which several can be considered cryptic. The crown node of Ehrharta is found to have originated 28 Ma, which substantially antedates previous age estimates. Comparisons of morphological evolutionary rates, as well as rates of nonsynonymous to synonymous sequence evolution (��), provide little evidence to support the hypothesis that non-ecological processes have been more important at high elevations. Both the low- and high-elevation clades show evidence of divergent ecological selection, with the Lowlands clade exhibiting high functional trait variance, and the high-elevation Setacea clade showing subtle ecological differentiation and accelerated rates of morphological evolution and �� relative to the rest of Cape Ehrharta. It is instead suggested that diversification in the Cape Ehrharta is triggered by a combination of intrinsic and extrinsic factors unique to each clade, thus corroborating a growing body of research arguing that it is simplistic to generalise radiations as ‘adaptive’ or ‘non-adaptive’. Instead, mountain radiations require an integrated approach to untangle the subtle interaction of geographic, ecological and biological factors that drive diversification.