Introduction. The aim was to analyse whether age at first drug offense predicts premature mortality and morbidity due to substance use and violence among adolescents and young adults. Methods. A prospective longitudinal register-linkage study based on a total population sample from Finland including individuals born between 1987 and 1992 and aged 15-25 years during follow-up in 2002-2017 (n = 386 435). Age-specific rates of deaths and health-care admissions (morbidity) during a 5-year follow-up were calculated from the first drug offense. Cox regression models were used to estimate differences in mortality and morbidity at ages 21-25. Results. Of all 15-to 20-year-olds, 1.4% (n = 5540) have had a police contact. The 5-year mortality rates (per 1000 person-years) among those with first drug offense at ages 15-16 was 2.92 [95% confidence interval (CI) 1.56-6.18], and 5.26 (CI 4.00-7.07) and 5.05 (CI 4.06-6.38) at ages 17-18, and 19-20, respectively. The rates of morbidity varied between 61.20 ). Both mortality and morbidity rates were over 10 times higher than among the general population. In models adjusted for family background, first police contact at an early age (15-16) did not increase the risk of mortality at ages 21-25 compared with first police contact at ages 17-18 (hazard ratio 1.55, CI 0.77-3.09) or 19-20 (hazard ratio 1.52, CI 0.78-2.98). The results were similar for morbidity. Discussion and Conclusions. Adolescents with drug-related police contacts have high risk of mortality and morbidity due to substance use and violence regardless of age of first contact.