Dennett's recent defense in this journal of the heterophenomenological method and its supposed advantages over Husserlian phenomenology is premised on his problematic account of the epistemological and ontological status of phenomenological states. By employing Husserl's philosophy of science to clarify the relationship between phenomenology and evidence and the implications of this relationship for the empirical identification of "real' conscious states, I argue that the naturalistic account of consciousness Dennett hopes for could be authoritative as a science only by virtue of the very phenomenological evidences Dennett's method consigns to the realm of fiction. Thus heterophenomenology, qua scientific method, is incoherent.