Over the last two decades, much progress in the prevention and early detection of first-episode psychosis has been achieved. And currently, the inclusion of a psychosis risk syndrome in DSM-V is discussed. But while differences between early-onset and adult-onset psychoses have frequently been pointed out, in the early detection of psychoses, developmental aspects are rarely considered, and results that have been derived from predominately adult samples, i.e., at-risk criteria are generally just transferred to children and adolescents. Emerging evidence from child and adolescent general population, birth cohort and clinical samples, however, question this unrestricted transferability of these results to younger age groups. This presentation will therefore outline the special aspects in the early detection of psychosis that should be considered in children and adolescents. The presentation will cover (i) evidence for the need of a validation of at-risk criteria in children and adolescents and (ii) evidence for a longer DUP in adolescent onset compared to adult onset psychosis and its clinical implications.