Results illustrate that the offending-victimization causal relationship is significant and stable over time, after dynamic protective/risk factors and persistent individual heterogeneity are controlled. This finding is consistent with previous research which suggests that adolescents' delinquent behavior is an important risk factor for victimization experience. The strength of the offending-victimization association remains stable during adolescence. That social bonding with conventional institutions mediated the offending-victimization linkage was not supported. Finally, there is support predicting that the offending-victimization association is partially spurious because of observed and unobserved population heterogeneity. The associations between offending and less serious violent victimization, such as being threatened and being jumped, are partially spurious owing to observed and unobserved stable individual characteristics; however the association between adolescent delinquent behavior and serious victimization, such as being shot or stabbed, remains almost the same when other risk factors and protective factors are taken into account. This suggests that the offending-victimization linkage, as well as the mechanism that leads to this linkage, appears to be victimization specific. Whereas most offending-victimization associations can be attributed to both offending and persistent individual propensities, the most violent victimization, such as being stabbed or shot, is largely the result of an increase in offending and a loss of connection to conventional institutions. Data were collected from the public use version of ADD Health, specifically, the two-wave longitudinal data that included 6,504 individuals at Wave 1 and 4,834 individuals at Wave 2. Tables, appendix, notes, and references