The present study assesses the relationship between family and educational disadvantage on self-reported offending, victimisation and violent youth group involvement in a Belgian medium-sized city. Many studies have focused on the relationship between family disadvantage (one-parent families, immigrant background) and educational disadvantage (vocational tracking, school failure) and violent youth group involvement, offending/victimisation in surveys. The present study primarily assesses to what extent social bonds (parental monitoring and the school social bond), deviant beliefs, low self-control and lifestyle risk are stable mediators of the relationship between family and educational disadvantage and self-reported offending, victimisation and troublesome youth group involvement among young adolescents. The results indicate that the lifestyle-exposure model, which was initially used to explain individual differences in victimisation is much better capable of explaining differences in selfreported offending and violent youth group involvement than victimisation. The implications for further studies are discussed.