We review theoretical explanations for in-kind transfers in light of the limited empirical evidence. After reviewing the traditional paternalistic arguments, we consider explanations based on imperfect information and self-targeting. We then discuss the large literature on in-kind programs as a way of improving the efficiency of the tax system and a range of other possible explanations including the "Samaritan's Dilemma", pecuniary effects, credit constraints, asymmetric information amongst agents, and political economy considerations. Our reading of the evidence suggests that paternalism and interdependent preferences are leading overall explanations for the existence of in-kind transfer programs, but that some of the other arguments may apply to specific cases. Political economy considerations must also be part of the story.