["\nAbstract\nIt has been argued that the predictors of all criminal career features are the same, and that childhood risk factors do not predict life‐course‐persistent offenders. Little is known about childhood predictors of the duration of criminal careers. The aim is to investigate childhood (aged 8–10 years) risk factors for criminal career duration, in comparison with childhood risk factors for other criminal career features. The Cambridge Study in Delinquent Development is a prospective longitudinal survey of 411 South London males from age 8, with conviction records up to age 61. Life‐course‐persistent (LCP) offenders were defined as those with a criminal career lasting at least 20 years. The strongest predictors of LCP offenders were harsh discipline, poor parental supervision, a convicted father and parental conflict. Childhood risk factors for LCP offenders and criminal career duration were different from childhood risk factors for the prevalence of offending (convicted versus unconvicted males). These results should be taken into account in developmental and life‐course criminology theories, risk assessment instruments and risk‐focused interventions.\n", "Criminal Behaviour and Mental Health, Volume 30, Issue 4, Page 159-171, August 2020. "]