Abstract Aim Little is known about the elevational gradient of plant endemism. It is mostly assumed that patterns are determined by topographical factors such as area of elevational belts and degree of habitat fragmentation, but comparative studies of different plant taxa along the same elevational gradient are lacking. The aim was to compare the elevational patterns of plant endemism of the entire flora and selected families and genera in a search for commonalities. Methods The elevational patterns of endemism for the entire Ecuadorean vascular plant flora, for twenty-seven selected families, and for twenty-four selected genera based on the Catalogue of Vascular Plants of Ecuador were analysed. Results Elevational patterns of endemism were non-random at all taxonomic levels but there was no common elevational pattern. Rather, the study groups showed a wide variety of independent patterns at all taxonomic levels. Most groups had hump-shaped patterns with maxima at different elevations and mostly at the same or at higher elevations than the maxima of species richness. The overall flora showed highest endemism in the narrowest and most fragmented elevational belts, presumably because of the consequent fragmentation of species populations. Main conclusions Patterns of endemism appear to be influenced both by taxon-specific ecological traits (e.g. life form, reproduction, dispersal, demography, spatial population structure, competitive ability) in their specific interaction with historical processes and by environmental factors such as topographical fragmentation. The degree to which these influences become visible along the elevational gradient are determined by which combination of species is analysed: for a given genus or family, taxon-specific traits dominate the patterns, for the entire flora taxon-specific patterns are blurred by averaging and the signal of topography emerges. Beyond the elevational gradient as such, this study shows that the frequently stated assumption that given biogeographical settings lead to similar patterns of endemism among different taxa is wrong.