Repeated victims of school bullying are at risk for maladjustment and could potentially benefit from selective intervention. However, selective intervention requires a practical method for accurately identifying members of the targeted group. We examined the extent to which the global victimization item from the Olweus Bully/Victim Questionnaire accurately identified children repeatedly victimized by peers. From a sample of 654 fourth-graders, we categorized 63 (9.7%) as repeated victims based on elevated reports of victimization from the same informant (self, teacher, peer) across two time points within an academic year. Logistic regression and ROC curve analyses examined the predictive utility of the recommended global item cutoff and more stringent cutoffs. Results indicated that the global victimization item offered limited utility as a means of identifying children categorized as repeated victims. We discuss possible reasons for the findings and potential implications for future research and practice.