Epistemologists disagree on the appropriate response to disagreement with a single peer. But there is wider acceptance that the greater the number of one’s peers against one, the more significance one should afford the disagreement. From this, we might suppose that an individual dissenter within a group of sufficient size and competence should accede to the collective judgement. Call this principle ‘Collective Superiority’. In contrast, discussions of disagreement within collective inquiry have highlighted its epistemic value. Following these discussions, a plausible principle of collective rationality is that justified collective judgements demand methods of inquiry that permit and preserve dissent. Call this ‘Epistemic Liberalism’. Taken together, ‘Collective Superiority’ and ‘Epistemic Liberalism’ allow for cases in which an individual dissenter should – qua individual rationality – accede to the collective judgement and at the same time should – qua collective rationality – stay steadfast in their dissent. Call this the ‘puzzle of individual dissent’. Whilst such conflicts between normative domains are familiar, I argue that this puzzle represents a genuine problem within the epistemology of disagreement. To resolve it we must reject one or the other or both of ‘Collective Superiority’ and ‘Epistemic Liberalism’. I conclude by considering the significance of rejecting ‘Collective Superiority’ within the epistemology of disagreement.