The relationship between victimization and offending is now well established. Research has shown, using comparative, cross-sectional, and longitudinal data, that victims are more likely to become offenders and vice-versa. To-date, a rigorous, quasi-experimental method has not yet been employed to estimate this effect of one variable on the other. This study uses propensity-score matching to investigate the robustness of the victim-offender overlap in a longitudinal, nationally-representative sample of adolescents. Results support the extant body of research on a victim-offender overlap.