The study examined competing hypotheses that attempt to explain the co-occurrence of wife and child abuse, specifically trait versus instrumental theories of aggression within families. It analyzed three commonly cited catalysts (unemployment, drinking, and life-stress events) for men's abuse of family members to determine whether they equally predict partner or child abuse. The study interviewed 363 women and children about spousal and parental abuse, and women were interviewed about sociodemographics and the stressors noted above, using the Conflict Tactics Scale. Families were oversampled for the presence of spousal violence. Logistic regressions revealed that heavy drinking and life stress events predicted men's abuse of their partners. These risk factors were unrelated to child abuse. Wife battering, however, placed children at heightened risk. Children of battered women stood a 42 percent chance of receiving escalated abuse from their fathers. The study proposes that men's abuse of children is in many instances instrumental in order to coerce or retaliate against women, using children as pawns in their abusive behavior. It suggests the need for measures to discourage this form of coercion and aggression to end the expression of patriarchal terrorism described by the women and children interviewed for this study. Tables, references