Should it be a constraint on metaphysical theories that we can believe them? I’ll suggest that it needn’t be; that “incredible” theories might nonetheless be serious contenders. I’ll take as my foil Lewis’ Maxim of Honesty: you ought not defend a theory that you yourself cannot believe. On the face of it, the Maxim functions like a kind of ‘common sense’ constraint on metaphysics, but gives belief a radically different role in theory choice than other more familiar Moorean appeals. In this talk I explore three varieties of incredible metaphysics, and argue that even Lewis’ minimal constraint oversteps. Along the way, I hope to cast some light on the vexed relationship between “ordinary belief” and metaphysical theorizing.