Approximately 500 students in 11 public school districts completed questionnaires in 1995. The final sample included 5,935 8th grade students representing 315 classrooms in 42 schools. Behavioral, demographic, and attitudinal traits were examined, as well as self-reported deviance. The relationship between variables representing family processes (attachment, supervision, and values socialization) and self-reported deviance was analyzed for both intact and non-intact families. Young people in non-intact households were more likely to be involved in both general and serious deviance than young people in intact households. However, when family process measures were incorporated in the analysis, involvement in deviant behavior was primarily due to adolescent values socialization, specifically adolescent neutralizations of behavior and perceptions of guilt. Parental monitoring behaviors were also related to lower levels of deviance, but findings regarding parental attachment were inconsistent across intact and non-intact family structures. Directions for future research are noted. An appendix includes index and scale information associated with the study. 43 references, 6 notes, and 3 tables