The problem how to ascertain the truth about the past is as old as history itself. But until the work of Louis Mink, no clear distinction was made between questions concerning the truth of statements on the past and questions concerning the truth of historical narratives as a whole. A narrative, Mink argues, is not simply a conjunction of statements on the past. Therefore its truth cannot be a function of the truth of its individual statements. The problem of narrative truth is according to him thus: although each statement (or set of statements) asserting a relation between events is subject to confirmation and disconfirmation, the combination of interrelations as established by the historical narrative is not, even though such combination of interrelations represents a real combination in past reality and is claimed to be true. As if to further complicate the problem, Mink maintains that history shares its form with fiction. Three and a half decades after Mink formulated the problem of narrative truth, it has not been dealt with in a satisfying manner. Mink does not solve nor dissolve the problem he posed. That task is taken up in this essay. It will move us away from the vocabulary of literary theory towards a pragmatist account of narrative truth. © Koninklijke Brill NV, Leiden, 2013.