["\nAbstract\nCriminal career duration has not been well investigated. There are very few longitudinal data sets that last long enough and enough subjects to investigate criminal career duration, and especially the characteristics and risk profiles of especially life course persistent offenders. The aim of the study was to describe the predictability of criminal career duration based on both official records and self‐reports of offending, and to put the results in the context of the Moffitt theory of “adolescent limited” and “life course persistent” offenders. The Pittsburgh Youth Study (n = 1517) is a seminal longitudinal study based on a community sample of high‐risk boys from the city of Pittsburgh. Data was used from the oldest sample of boys in the PYS (N = 506). The participants were first assessed on average at age 12, and data was used up until age 36 for self‐reported offending, and age 40 for police charges. The analyses were conducted on moderate and serious violence and moderate and serious theft. Career duration was based on self‐reports and official charges in combination. The results show the extent to which commonly accepted and well validated risk factors predict criminal career duration with a special focus on individuals showing high rate/persistent offending in adolescence. Results show more pathways of delinquent development than have previously been described by Moffitt, (Psychological Review, 1993, 100(4), 674–701). Results also show that there is limited predictability of delinquency and offending over time. In addition, results show that only a small number of risk factors distinguish high and low rate adolescent offenders who become continuous offenders. Examples are peer factors. Implications for policy making and intervention science are discussed.\n", "Criminal Behaviour and Mental Health, Volume 30, Issue 4, Page 172-182, August 2020. "]