We examined how family and child risk factors jointly affected stability and change in externalizing behavior over time in a prospective study of eventual alcohol use disorder. Study participants were community-recruited alcoholic and control families, and their initially preschool-aged male and female children (N = 335). Family risk varied as a function of both parental alcoholism (ALC) and antisocial personality disorder (ASPD) and was evaluated for both parents. Child risk was characterized by a set of risky temperament attributes pertaining to high activity, high reactivity, and low attention span. Externalizing behavior was used as the proxy indicator for later alcohol problems. For children in the high family risk group (involving current ALC in both parents or current ALC + ASPD comorbidity or both), child risk when children were 3–5 years old (Wave 1) directly predicted externalizing behavior when children were 6–8 years old (Wave 2), even when Wave 1 child risk was controlled for. In addition, parents' negative interaction with children at Wave 1 mediated the effect of child risky temperament on Wave 2 externalizing behavior. No such pattern was observed in the low family risk group, where only autostability effects were predictive of outcomes at Wave 2. The importance of nesting structure as an ingredient in the epigenesis of risk was discussed. Its particular relevance in understanding the process of risk transmission among offspring from antisocial alcoholic families was emphasized.