We suggest that in some instances the apparent logical inconsistency of moral hypocrisy stems from different evaluations of a weak argument, rather than dishonesty per se. Extending Corner, Hahn, and Oaksfords (2006) analysis of slippery slope arguments, we propose that inferences of hypocrisy depend on perceived similarity of actions to previous standards. In Experiment 1, dissimilar actions were rated as less hypocritical than their similar counterparts. If observers are choosing between competing theories (i.e., hypocrisy or legitimate dissimilarity), evidence of self-serving motives will positively support inferences of hypocrisy independent of changes in similarity. In Experiment 2, we manipulated potential self-serving interests that an action would produce while keeping similarity between cases identical. Actions that would result in a beneficial outcome for the actor were seen as more hypocritical than their non-self-serving counterparts. These results support the possibility that Bayesian analyses of weak arguments have implications for assessing moral reasoning.