Although literature on traditional bullying is abundant, a limited body of sound empirical research exists regarding its newest form: cyber bullying. The sample comprised Australian secondary students (N = 803) and aimed to identify the underlying structure of cyber bullying, and differences in traditional and cyber bullying behaviors across gender and grade. Reliability analyses, confirmatory factor analyses, and factorial invariance testing demonstrated that the newly extended measure of traditional and cyber bullying was psychometrically sound. Multiple-Indicators-Multiple-Causes models demonstrated gender, grade, and gender by grade interaction effects for traditional and cyber forms of bullying and being bullied. Findings were interpreted in the context of bullying theory. Moreover, potential limitations of the investigation and implications for theory, research, and practice were discussed. (Contains 11 tables and 3 figures.)