Purpose: This study offers a partial test of Moffitt's (1993) hypothesis that some adolescent-limited offenders may be ensnared by the consequences of their delinquency in a way that continues their offending. Methods: Cambridge Study in Delinquent Development data is used along with a multi-step analytic procedure including (a) group-based trajectory modeling to identify variability in offending and (b) propensity score matching to estimate the impact of one specific snare, heavy drinking, on criminal convictions occurring in early adulthood. Results: Results provide some support for Moffitt's hypothesis; adolescent-limited offenders who engaged in heavy drinking at age 18 were more likely to be convicted during early adulthood. Conclusions: This analysis offered a stronger test of the snares hypothesis by first, separating the data into groups of offenders who followed similar offending patterns, and second, matching the cases on developmental and psychosocial characteristics in order to assess the snare's role on later offending. Findings call attention to the effect of problem adolescent behavior on later adult outcomes.