Imagine a game of chess. As the players think about their next moves, they consider the outcome each action would have on the board, as well as the opponent’s likely reply. The players’ knowledge of the board and the rules constitutes an internal model of chess, a knowledge structure that links actions to their likely outcomes. The process of using such an ‘action–outcome’ model to inform behavior is defined within reinforcement learning theory as the act of planning. Planning, so defined, has been an object of scientific investigation for many decades, and this research has generated important insights into the planning abilities of both humans and other animals.