This 3-year follow-up study compares background variables, extent of criminality and criminal recidivism in the form of all court convictions, the use of inpatient care, and number of early deaths in Swedish institutionalized adolescents (N = 100) with comorbid substance use disorders (SUD) and Attention-Deficit/Hyperactivity Disorder (ADHD) (n = 25) versus those with SUD but no ADHD (n = 30), and those without SUD (n = 45). In addition it aims to identify whether potential risk factors related to these groups are associated with persistence in violent criminality. Results showed almost no significant differences between the three diagnostic groups, but the SUD plus ADHD group displayed a somewhat more negative outcome with regard to criminality, and the non-SUD group stood out with very few drug related treatment episodes. However, the rate of criminal recidivism was strikingly high in all three groups, and the use of inpatient care as well as the number of untimely deaths recorded in the study population was dramatically increased compared to a age matched general population group. Finally, age at first conviction emerged as the only significant predictor of persistence in violent criminality with an AUC of .69 (CI (95%) .54–.84, p = .02). Regardless of whether SUD, with or without ADHD, is at hand or not, institutionalized adolescents describe a negative course with extensive criminality and frequent episodes of inpatient treatment, and thus requires a more effective treatment than present youth institutions seem to offer today. However, the few differences found between the three groups, do give some support that those with comorbid SUD and ADHD have the worst prognosis with regard to criminality, health, and untimely death, and as such are in need of even more extensive treatment interventions.