ABSTRACT The human brain is a highly complex biological tissue organized into hundreds of regions composed of a myriad of cell types with distinct molecular, morphological, and physiological properties. These cells and their associated circuits underlie our mental abilities and, when dysfunctional, lead to neurological and psychiatric disorders. Consequently, developing an atlas of these cell types and how they differ from one another is essential for understanding the biological processes underlying human brain development and function. In addition, because both the general cytoarchitecture of the brain and its constituent cells are generally conserved across primates, knowledge of how these cell types differ between species is also essential for understanding uniquely human aspects of cognition and behavior. We therefore propose to use single cell transcriptomics to rigorously define cell types and states in ten brain regions involved in higher cognition and behavior and across five key developmental timepoints of human, chimpanzee (one of our closest extant relatives), rhesus macaque (the most commonly studied Old World non-human primate), and common marmoset (an emerging New World non- human primate model system). By generating, analyzing, and integrating these data with existing unpublished datasets, we propose to develop a human and non-human primate cell atlas and identify shared and divergent molecular and cellular features across species, regions, and ages. Furthermore, we will validate key aspects of this atlas, including molecular signatures, and create a data visualization and dissemination portal.