Cognitive neuroscientists frequently talk about the brain representing the world. Some philosophers claim that this is a confusion. This paper argues that there is no confusion, and outlines one thing that ‘the brain represents the world’ might mean, using the notion of a model derived from the philosophy of science. This description is then extended to make apply to propositional attitude attributions. A number of problems about propositional attitude attributions can be solved or dissolved by treating propositional attitudes as models.