This paper offers a new interpretation of Peirce’s “The Fixation of Belief,” one that pays special attention to the notion of what it is for a belief to be fixed or stable. Drawing on Louis Loeb’s stability-centered readings of the epistemologies of Descartes and Hume, I suggest that Peirce’s argument does not deploy a Cartesian notion of permanently stable belief. It instead relies on a comparatively modest Humean notion, according to which a fixed belief is stable for a given inquirer at a given time. A belief is settled if it resists being undermined too readily by experience and reflection. I try to show how such a reading allows Peirce to object to unscientific methods of fixing belief purely on grounds of their ineffectiveness. I try also to show how Peirce establishes the superior effectiveness of the method of science at delivering stable beliefs in the relevant sense.