We investigate the labour market effects of incarcerating children. Using linked administrative data to track outcomes for English schoolchildren, we estimate an econometric model of transitions between education, custody, employment and NEET (not in employment, education or training), along with earnings for those starting work. We allow outcomes to vary according to the individual’s state in the preceding spell and, by controlling for personal characteristics and unobserved heterogeneity, interpret such variation as capturing causal impacts. For males, the main effect of incarceration is a reduction of more than 10% in the probability of employment. For females, there is no overall impact on employment but, for those entering work, wages are reduced by 25%. These negative impacts suggest roles for policy in deterring delinquency, finding alternatives to custody, rehabilitating those incarcerated and supporting resettlement on release. Appropriate labour market policy may differ by gender, with males needing help to overcome employer discrimination and females needing encouragement to achieve better-paid work.