Five studies examined Kahneman and Miller's (1986) hypothesis that events become more "normal" and generate weaker reactions the more strongly they evoke representations of similar events. In each study, Ss were presented with 1 of 2 versions of a scenario that described the occurrence of an improbable event. The scenarios equated the a priori probability of the target event, but manipulated the ease of mentally simulating the event by varying the absolute number of similar events in the population. Depending on the study, Ss were asked to indicate whether they thought the event was due to chance as opposed to (a) an illegitimate action on the part of the benefited protagonist, or (b) the intentional or unintentional misrepresenation of the probability of the event. As predicted, the fewer ways the events could have occurred by chance, the less inclined Ss were to assume that the low-probability event occurred by chance. The implications of these findings for impression-management dynamics and stereotype revision are discussed.