Data were obtained from the 1996, 1997, and 1998 National Crime Victimization Surveys (NCVS), which is an annual survey of approximately 50,000 households in the United States. Youth (ages 12-17) were placed in the domestic-violence-household group when at least one of the adult members experienced a violent victimization in a given year at the hands of a domestic partner. Youth placed in the non-domestic-violence group were those in which one adult member of the household experienced violent victimization in a given year at the hands of someone not living in the household; and youth in the no-crime group were in a household in which there was no violent or nonviolent victimization of any adult member in a given year. Data analysis used the STATA Release 6, because this software has commands that account for complex sample design in survey analyses. Logistic regression was used to identify the variables related to violent victimization risk among the three groups of youth. Results were determined for all youth, girls only, and boys only. The findings showed a generally high victimization risk for youth who lived in households with a victimized adult. The risk was elevated for youth from households with adult victims of both domestic and nondomestic violence. Girls who lived in households with an adult victim of domestic violence were apparently at higher risk for crime victimization. Some of the added risk was from family members. These findings reinforce the need to target crime prevention and detection efforts toward youth in households in which an adult has been victimized. 6 tables, 1 figure, and 25 references