The model included one exogenous variable (observing marital violence during childhood) and five endogenous variables (the participant's attitude regarding the acceptability of marital violence, his level of marital stress, his level of sex-role egalitarianism, his level of alcoholism, and his level of self-esteem). The model was tested using data from 115 men, of whom 44 were in treatment for male violence and 71 were in alcoholism treatment. Results of the path analytic procedure suggested that sex-role egalitarianism and approval of marital violence both have direct effects on the use of severe marital violence. At the same time, sex-role egalitarianism and the observation of marital violence as a child have indirect effects by several paths. Egalitarianism has an indirect negative effect on the use of severe violence, which depends on approval of marital violence. The observation of marital violence has a negative effect on self-esteem, influencing marital stress and the level of alcoholism, both of which have an effect on the approval of marital violence. The observation of violence as a child also has a direct effect on the approval of violence and a negative effect on sex-role egalitarianism. Figure, table, and 59 references (Author abstract modified)