A rich history of work in cognitive development focuses on children’s emerging understanding of various conceptual domains, such as inanimate objects, human-made artifacts, and animal kinds. However, one of the most striking features of human societies is the presence of distinctly social kinds, whose relevance emerges from a complex network of distributed beliefs and practices. Sometimes referred to as institutional kinds, salient examples include money, borders, and diplomas. These entities are real and causally efficacious, but they operate in quite different ways than do “standard” artifacts such as hammers and walls. I will present recent work focusing on how children come to understand these sorts of entities, as well as a parallel case of institutional social roles such as judges and presidents, which can again be contrasted with “standard” social roles such as carpenters and doctors. This is new work that I’m eager to get feedback on, especially with respect to a seeming puzzle: Why does it take children so long to grasp the causal structure of institutional kinds, despite their much earlier understanding of other aspects of social reality such as rules and norms?