In this chapter I show that one of the paramount problems of epistemology, Gettier’s problem, is solved if one assumes that knowledge, truth and justification are relative to representations, theories, or, if one prefers, to coherent sets of hypotheses. This idea comes from practice first: in many somewhat complex cognitive contexts, people use the words ‘know’, ‘true’ and ‘justified’ in exactly that way: by (explicitly or implicitly) referring to some set of hypotheses or theory. Phrased slightly more precisely: these persons’ conversations, when they use these words, can be made fully coherent if one assumes that there is an (implicit) reference to a theory (understood in a general sense, denoted by ‘theory*’). Notice how well this claim agrees with the MCT-principle, implying, among other things, that coherent answers to non-trivial questions—knowledge—are given only by whole theories. I show that this relativistic account strongly resonates with Quine’s acclaimed holism, and solves other well-known problems of epistemology. If we combine the results of this chapter with other theses of this book, and acknowledge that it has proven difficult to find other solutions to Gettier’s problem, we arrive at the conclusion that knowledge, truth and justification must be relative to theories in a fundamental sense. Thus, representation-relativity is the core principle of my epistemology. All in all, the results obtained in this chapter lend credit to a type of moderate epistemic relativism, call it ‘theory relativism’ (not, of course, in contradiction with realism), in which the notion of ‘absolute truth’ can be recovered as a natural limit case.