The study found that family dissolution was strongly associated with criminal behavior by the affected children, although the strength of this association was significantly reduced when other variables that captured social and economic conditions were controlled; still, family dissolution remained a strong predictor of a child's criminal behavior. The findings generally agree with those of previous similar studies in terms of the importance of family functioning, conflicts, and emotional stress. Family characteristics indirectly affected other variables, such as the resources and opportunities available to children in the course of their development. Crime statistics (criminal charges instead of convictions) were collected on the birth cohort from ages 10 to 22 years old. "Family type" referred to whether a person's parents lived together. "Family dissolution" was operationalized as families in which parents were divorced or had decided to discontinue cohabiting if unmarried. Other variables measured were parent's education, teen pregnancy, father and mother's income from work, father's and mother's social benefits, father or mother charged for a drug offense, and father or mother imprisoned. Data on these variables were collected regularly for the years 1992 to 2004 (ages 10 to 22). 3 tables, 1 figure, 36 references, and appended mathematical equations used in data analysis