Why some animals have big brains and others do not has intrigued scholars for millennia. Yet, the taxonomic scope of brain size research is limited to a few mammal lineages. Here, we present a brain size dataset compiled from the literature for 1,552 species with representation from 28 extant taxonomic orders. The brain–body size allometry across all mammals is (Brain) = –1.26 (Body)^0.75. This relationship shows strong phylogenetic signal. Thus, we conducted additional allometries using median species values for each order, family, and genus to ensure evolutionary independence. Slopes from these analyses at different taxonomic levels all approximate ∼0.75 scaling. Why brain size scales to the 3/4 power to body size across mammals is to our knowledge unknown. Slopes within taxonomic orders, exhibiting smaller size ranges, are generally shallower than 0.75 and range from 0.24 to 0.81 with a median slope of 0.64. Published data on brain size are lacking for the majority of extant mammals (> 70% of species) with strong bias in representation from Primates, Carnivora, Perissodactyla, and Australidelphian marsupials (orders Dasyuromorphia, Diprotodontia, Peramelemorphia). Several orders are particularly underrepresented. For example, data on brain size are available for less than 20% of species in each of the following speciose lineages: Soricomorpha, Rodentia, Lagomorpha, Didelphimorphia, and Scandentia. Use of museum collections can decrease the current taxonomic bias in mammal brain size data and tests of hypothesis.