Conservation organizations are collecting large-scale data regarding distribution and threats to vertebrate taxa. These data sets will enable planners to systematically identify large-scale conservation priorities; however, they will cover only a tiny proportion of living organisms. Therefore, it is essential to investigate to what extent the areas selected for conservation actions can provide protection for other species. We analyzed the umbrella effect between amphibians and mammals across mainland Africa. We built habitat suitability models within the geographic ranges of 1654 species, based on data collected in the framework of the World Conservation Union (IUCN) Global Amphibian Assessment and IUCN Global Mammal Assessment. We applied systematic reserve selection algorithms to two sets of estimators of the area of occupancy of amphibians and mammals (geographic ranges and estimated suitable areas) and thus selected four reserve systems. We then quantified the protection that each of the four systems provided for amphibians and mammals. Reserves selected for amphibians and mammals were comparable in area, with the former concentrated in the Afrotropical region and the latter more evenly dispersed. Mammal reserves left fewer gaps in species coverage among amphibians than the reverse, but amphibian reserves included a larger proportion of each mammal's area of occupancy than the reverse. For both taxa, setting reserves to include estimated suitable areas instead of ranges resulted in the clustering of reserves in the tropics. Furthermore, it efficiently protected hidden gaps (species with unsuitable portions of their range inside protected areas) in the other taxon and included a higher proportion of the area of occupancy of the other taxon. Overall, amphibians and mammals in Africa acted as an umbrella for a high proportion of species in the other taxon. Focusing on estimated suitable areas instead of ranges improved the umbrella effect of both taxa.