According to the perceptual model, our knowledge of others' minds is a form of perceptual knowledge. We know, for example, that Jones is angry because we can literally see that he is. In this essay, I argue that mental states do not have the kind of distinctive looks that could sufficiently justify perceptual knowledge of others' mentality. I present a puzzle that can arise with respect to mental states that I claim does not arise for non-mental properties like being an apple and argue that this is explained by the fact that the looks of non-mental properties adhere to a certain explanatory principle that does not hold for mental states. This shows, I argue, that, even if we think mental states do have looks, these cannot offer sufficient grounds for perceptual knowledge of others' minds. In the final section of the essay, I suggest an alternative way of thinking about our knowledge of others' minds and about the sorts of looks or appearances that might be associated with mental states. Many people have the sense that we simply cannot know about the mental states of other individuals in the same way we know about ordinary material objects. Nevertheless, a growing number of philosophers have come to endorse the surprising thesis that our knowledge of others' minds is a form of perceptual knowledge. They claim, for example, that we know that Jones is angry because we can literally see that she is angry (e.g., Cassam, 2007;