Given the rapid increase in studies of bullying and peer harassment among youth, it becomes important to understand just what is being researched. This study explored whether the themes that emerged from "children's" definitions of bullying were consistent with theoretical and methodological operationalizations within the research literature, and whether the provision of a definition when administering bullying experience items would lead to different prevalence rates in reported victimization and bullying. Students aged 8-18 (N = 1767) were randomly assigned to one of two conditions. In the first condition, students were provided with a standard bullying definition; in the second condition, students provided their own definition of bullying. Results indicated that students' definitions of bullying rarely included the three prominent definitional criteria typically endorsed by researchers: intentionality (1.7%), repetition (6%), and power imbalance (26%), although almost all students (92%) did emphasize negative behaviors in their definition. Younger children made more mention of physical aggression, general harassing behaviors, and verbal aggression in their definitions, whereas the theme of relational aggression was most prominent in the middle years and reported more by girls than boys. Finally, students who were given a definition of bullying reported being victimized less than students not provided with a definition. As well, boys who were given a definition of bullying tended to report higher levels of bullying than those not given a definition (marginal effect). (Contains 3 tables and 3 figures.)