According to moral error theorists, moral discourse is guilty of a systematic error, and so no positive first-order moral claims can be true. Suppose that a moral error theory turns out to be correct. In that case, what ought we to do with our moral practices? Moral fictionalists recommend that we preserve moral discourse in the spirit of a useful fiction—i.e., that we make-believe and pretend to assert moral propositions (Joyce 2001; Nolan, Restall, & West 2005). Fictionalists promise us the best of both worlds: we avoid incurring the epistemic cost of holding onto our false moral beliefs, and we preserve the practical goods associated with moral practice. However, I argue that characteristic features of make-belief as opposed to belief suggest that fictionalist attitudes won’t be a stable enough or strong enough basis for providing the practical benefits that we want here.