We each seem to have a non-inferential way of knowing certain explanations of our own actions and mental states, a way by which others cannot know such explanations. For example, I seem to have a non-inferential way of knowing that I am waiting in my office because I am expecting a phone call. But things are not always as they seem. Richard E. Nisbett and Timothy DeCamp Wilson famously set out to test experimentally the claim that we have a non-inferential way of knowing such explanations. They argued that the kind of differences we would expect between ourselves and others if this claim were true do not exist. They further argued that the best explanation of this fact is that we must be arriving at explanations of our own actions and mental states by inference, just as others are. In this paper, I explicate and evaluate Nisbett and Wilson’s argument. I argue that Nisbett and Wilson’s claim that we arrive at such explanations of our own actions and mental states by inference is not adequately supported by their findings.