In recent years a new way to think about the brain and the mind is in terms of Predictive Coding or Predictive Processing. On the contrary to traditional ideas of cognition and perception, this innovative framework claims that the brain a prediction machine that creates models of the world and tests these prediction models against the incoming sensory input. This view does not only change our metaphysical assumption about how the mind works, it also changes the epistemic setting. In this context, one interesting question arises, namely, how does this influence our philosophical accounts of self-knowledge. In this talk, I will examine accounts of self-knowledge and the assumption that self-knowledge is privileged within the Predictive Processing framework. I will suggest that Predictive Processing has some constraints on how we can elaborate self-knowledge accounts. At the same time, however, it is not clear, whether this framework touches the most fundamental questions.