Activity period plays a central role in studies of primate origins and adaptations, yet fundamental questions remain concerning the evolutionary history of primate activity period. Lemurs are of particular interest because they display marked variation in activity period, with some species exhibiting completely nocturnal or diurnal lifestyles, and others distributing activity throughout the 24-h cycle (i.e., cathemerality). Some lines of evidence suggest that cathemerality in lemurs is a recent and transient evolutionary state (i.e., the evolutionary disequilibrium hypothesis), while other studies indicate that cathemerality is a stable evolutionary strategy with a more ancient history. Debate also surrounds activity period in early primate evolution, with some recent studies casting doubt on the traditional hypothesis that basal primates were nocturnal. Here, we used Bayesian phylogenetic methods to reconstruct activity period at key points in primate evolution. Counter to the evolutionary disequilibrium hypothesis, the most recent common ancestor of Eulemur was reconstructed as cathemeral at ∼9-13 million years ago, indicating that cathemerality in lemurs is a stable evolutionary strategy. We found strong evidence favoring a nocturnal ancestor for all primates, strepsirrhines and lemurs, which adds to previous findings based on parsimony by providing quantitative support for these reconstructions. Reconstructions for the haplorrhine ancestor were more equivocal, but diurnality was favored for simian primates. We discuss the implications of our models for the evolutionary disequilibrium hypothesis, and we identify avenues for future research that would provide new insights into the evolution of cathemerality in lemurs.