The current research presents a new type of social context effect on attitude certainty. It is proposed that when people receive persuasive messages, they appraise their attitudes not only in terms of whether they are shared or not shared by others, but also in terms of whether they are based on similar or dissimilar assessments of the information presented. In two experiments, participants were presented with persuasive messages. In Experiment 1, they were induced to perceive that they responded favorably (persuasion) or unfavorably (resistance) to the message arguments. In Experiment 2, they were allowed to vary in their actual message responses. In both experiments, message response similarity—the degree to which people perceived that their evaluations of persuasive arguments were shared or unshared by others—moderated the classic effect of attitude similarity on attitude certainty. In particular, attitude similarity only affected attitude certainty under conditions of message response similarity. When message responses were believed to be dissimilar, attitude similarity had no effect on attitude certainty.