In this article I examine the ground and validity of Peirce's claim that 'belief has no place in science'. Contrary to the general view, such a claim should not be understood as merely an overreaction to William James' thesis that there can be legitimate non-evidential reasons to believe. For Peirce, believing that something is the case implies, at least when believing takes a certain form, a kind of dogmatism incompatible with the proper conduct of inquiry towards truth. In this paper, I examine two ways in which Peirce argues for the 'no belief in science' thesis. I first discuss 'his claim that belief is incompatible with the 'Will to Learn'. Peirce argues that believing that p in such a way that one does not have any real doubts about whether p implies that one has a dogmatic attitude vis-à-vis possible future evidence that not- p; I claim that this anticipates the line of reasoning that supports Kripke's 'paradox of dogmatism'. After having indicated how they can both be resisted, I examine a second way- which seems to have been overlooked in Peirce scholarship-in which the founder of pragmatism argues for the 'no belief in science' thesis. Peirce often relates this thesis to his view of abduction and the nature of explanatory hypotheses: the conclusion of an abductive inference is not, and should not be, the belief that a given explanatory hypothesis H is true, probably true, or likely to be true, but rather that H is such that it is a possible explanation of fact F.