Hotspots, which have played a central role in the selection of sites for reserves, require careful rethinking. We carried out a global examination of distributions of all nonmarine mammals to determine patterns of species richness, endemism, and endangerment, and to evaluate the degree of congruence among hotspots of these three measures of diversity in mammals. We then compare congruence of hotspots in two animal groups (mammals and birds) to assess the generality of these patterns. We defined hotspots as the richest 2.5% of cells in a global equal-area grid comparable to 1 degrees latitude x 1 degrees longitude. Hotspots of species richness, "endemism," and extinction threat were noncongruent. Only 1% of cells and 16% of species were common to the three types of mammalian hotspots. Congruence increased with increases in both the geographic scope of the analysis and the percentage of cells defined as being hotspots. The within-mammal hotspot noncongruence was similar to the pattern recently found for birds. Thus, assigning global conservation priorities based on hotspots is at best a limited strategy.