Abstract Background The juvenile justice system in the USA adjudicates over seven hundred thousand youth in the USA annually with significant behavioral offenses. This study aimed to test the effect of juvenile justice involvement on adult criminal outcomes. Methods Analyses were based on a prospective, population-based study of 1420 children followed up to eight times during childhood (ages 9–16; 6674 observations) about juvenile justice involvement in the late 1990 and early 2000s. Participants were followed up years later to assess adult criminality, using self-report and official records. A propensity score (i.e. inverse probability) weighting approach was used that approximated an experimental design by balancing potentially confounding characteristics between children with v. without juvenile justice involvement. Results Between-groups differences on variables that elicit a juvenile justice referral (e.g. violence, property offenses, status offenses, and substance misuse) were attenuated after applying propensity-based inverse probability weights. Participants with a history of juvenile justice involvement were more likely to have later official and violent felony charges, and to self-report police contact and spending time in jail (ORs from 2.5 to 3.3). Residential juvenile justice involvement was associated with the highest risk of both, later official criminal records and self-reported criminality (ORs from 5.1 to 14.5). Sensitivity analyses suggest that our findings are likely robust to potential unobserved confounders. Conclusions Juvenile justice involvement was associated with increased risk of adult criminality, with residential services associated with highest risk. Juvenile justice involvement may catalyze rather than deter from adult offending.