This paper argues that (1) if we want to explain why rationality is normative in the epistemic domain, we should reject epistemic consequentialism, and (2) a version of epistemic non-consequentialism I call Epistemic Kantianism provides the best explanation.  According to the Epistemic Kantian’s story, rationality is normative in the epistemic domain because complying with apparent epistemic reasons constitutes *respect for the truth*, and there is stringent objective reason, not merely apparent reason, to manifest such respect.  After introducing the topic in §1, I explain in §2 why epistemic consequentialism undermines the normativity of epistemic rationality.  In §3, I rehearse arguments against a thesis about derivative value I call Instrumentalism that is widely rejected outside epistemology though still popular among epistemologists, and show that epistemic consequentialism is either false or unmotivated if Instrumentalism is false.  In §4, I show how Epistemic Kantians can explain the normativity of epistemic rationality.  Before concluding in §6, I ask in §5 how strong a version of non-consequentialism is needed to explain the normativity of epistemic rationality, and suggest that a full vindication does indeed require Epistemic Kantianism.