Using data on males in the Cambridge Study, the current study found that experiencing a broken family up to the age of 14 predicted a violent criminal conviction by age 50. The likelihood of a conviction for a violent offense was moderated, however, by the variables of having a nervous disposition and harsh punishment. The effect of a later conviction for violence was mediated by several variables, with the most important being self-reported violence and hyperactivity, both measured at age 14. It may be that the stress associated with family breakdown leading to hyperactivity/restlessness and violent convictions may be causal; whereas, the family breakdown leading to self-reported violence and violent conviction may reflect a continuation of an underlying tendency toward violence. With respect to General Strain Theory, the findings provide some evidence that broken homes, a type of strain, have direct and indirect effects on offending, a result consistent with General Strain Theory. Additional evidence suggests that the relationship is also moderated by several variables and that the broken-home effect is heightened in the presence of some risks. This finding is consistent with Agnew's (2006) suggestion that researchers should consider the conditions under which strain may lead to more or less offending. Suggestions are offered for ways that parents and community-based social service agencies can mitigate the criminogenic effects that can stem from broken homes. 4 tables, 1 figure, 8 notes, and 89 references