The present study used the 2009-2010 sample of the Health Behavior in School-Aged Children Survey (N = 12,642) to investigate the incremental validity of four competing bullying involvement classification schemas, which differ as a function of relative rates of endorsing victimization and perpetration behaviors at school: the standard four-group model, an alternative four-group model, a nine-group model, and a sixteen-group model. The prevalence rates and associations with concurrent mental health and school functioning outcomes were explored for each of the four bullying involvement classification schemas, which were derived using a categorical analytic approach that relied on the same two self-report items--one targeting victimization and one targeting perpetration--in the HBSC self-report survey. Results indicated substantially different prevalence rates among the four schemas and that schemas which differentiated between higher frequencies of perpetration and victimization involvement were associated with greater odds of being at risk of psychological distress, psychological well-being, academic performance, and attitudes toward school. Findings also suggested that a continuous approach to deriving bullying involvement classifications may be useful, yet preliminary post hoc analyses indicated that this approach would yield redundant results given the parameters of the present study. Implications for school mental health practice and future research are discussed.