While some researchers suggest that higher taxon richness be used for setting conservation priorities, others claim that even data on ‘accepted species’ are inadequate, and that documentation of the finest patterns of differentiation is needed. The significance of taxonomic resolution was assessed using distributional data for 798 Andean bird species, which were recorded in 15′ × 15′ grid cells. This dataset was used because of the fine spatial resolution and the complex biogeographic patterns in this region. The primary database served as a template for creating three new databases, namely for genera, broadly defined biospecies and phylogenetic species. With this spatial resolution, the taxic richness ‘hotspots’ were roughly the same on all taxonomic levels, but the ratio between generic and species richness increased towards the biologically-poor high plateaus. Richness ‘hotspots’ covered only a small fraction of the endangered species. ‘Hotspots of endemism’ provided better guidance, and complementarity was decidedly best, but with generic data neither of these approaches were effective. On the other hand, the recognition of phylogenetic species revealed few new areas of endemism which were not also apparent using ‘accepted’ species. The analyses drew attention to aggregates of endemic species which are often sited in areas immediately adjacent to areas with dense human populations, possibly because traditional human settlements relied on ecoclimatic conditions which also determined the peaks of endemism. Since data are available for some taxonomic groups for detailed analysis of species data, there is little reason to base the conservation planning on coarser surrogates.