Abstract There has been extensive debate over whether we can have phenomenal knowledge in the case of epiphenomenalism. This article aims to bring that debate to a close. I first develop a refined causal account of knowledge—one that is modest enough to avoid various putative problems, yet sufficiently robust to undermine the epiphenomenalist position. I then consider and reject several possible responses that an epiphenomenalist might offer. The discussion thus shows that epiphenomenalist accounts of phenomenal knowledge—such as the third‐factor model and the constitutional model—are ultimately indefensible.