Using both quantitative and qualitative empirical research methods based on Social Disintegration Theory (SDT), this thesis, with 1936 stratified random cases of incarcerated offenders from 4 prisons in Zhejiang Province, studies the correlations of positional recognition, moral recognition, emotional recognition and voluntary acceptance of norms of non-violence on incarcerated offenders of new-generation migrant workers. It is hypothesized that incarcerated offenders of new-generation migrant workers will report more offence than incarcerated offenders of first-generation migrant workers, and that there will be cross-generation differences in sentenced violent crime and recidivism. Finally, the three dimensional independent variables will impact the cross-generation differences. The study first carries out a comparative analysis on the differences of all groups. The differences of the three independent variables are compared separately. In addition, special focus is also placed on the differences between first-generation and new-generation migrant workers in dependent variables and independent variables. Secondly, the thesis provides a correlation analysis on the independent variables and dependent variables of the entire cases and each group separately. Lastly, regression analysis is conducted for all groups together for each group. Support for the main hypothesis that migrant workers become the major crime group and new-generation migrant workers are higher than other groups in crime rate and violent crime rate was obtained. From positional recognition, variables of education, family, and income directly affect the occurrence of crime, and it is applicable to all groups. Variables of moral recognition are significantly correlated with offence and recidivism, but show little correlation with sentenced violent crime and sentence length. In terms of emotional recognition, family cohesion, family iii education, and parents’ deviant behaviors can generate some influence on the offence and violence of migrant workers. Furthermore, some interesting findings reveal that the groups of farmers and migrant workers perceive the strongest discrimination. As to migrant workers, institutional discrimination does not lead to cross-generational differences, and is not correlated with offence or violence, but non-institutional discrimination is correlated with offence and violence. Meanwhile, migrant workers with left-behind-home experience report more offence than those without the experience, but left-behind-home experience is not correlated with their sentenced violent crime.