The correlates of mammalian sleep have been investigated previously in at least eight comparative studies in an effort to illuminate the functions of sleep. However, all of these univariate analyses treated each species, or taxonomic Family, as a statistically independent unit, which is invalid due to the phylogenetic relationships among species. Here, we reassess these influential correlates of mammalian sleep using the formal phylogenetic framework of independent contrasts. After controlling for phylogeny using this procedure, the interpretation of many of the correlates changed. For instance, and contrary to previous studies, we found interspecific support for a neurophysiological role for rapid-eye-movement sleep, such as memory consolidation. Also in contrast to previous studies, we did not find comparative support for an energy conservation function for slow-wave sleep. Thus, the incorporation of a phylogenetic control into comparative analyses of sleep yields meaningful differences that affect our understanding of why we sleep.