How should banks be regulated? Must governments tightly regulate banks to prevent financial panics, or is little or no regulation best? Can private banks be trusted to issue paper money or must this activity be a government monopoly? Theory can help answer these questions, but increasingly in recent years economists have turned to the natural experiments of history to find out how well free banking systems, or more accurately lightly regulated banking systems, have worked in practice. We now have numerous studies of lightly regulated banking in Scotland, the United States, Canada, and many other countries. As usual, research has produced new questions and heated controversies. The resulting ruckus tends to obscure the areas in which research has produced a consensus. Here we try to separate the areas where there is a consensus from areas where research is still in its early stages.