DATA FROM A NATIONALLY REPRESENTATIVE SAMPLE OF 2,143 AMERICAN COUPLES SHOW THAT RATES OF DOMESTIC VIOLENCE ARE RELATED TO ATTITUDES ABOUT VIOLENCE AND TO SOCIAL STRUCTURAL VARIABLES SUCH AS INCOME, EMPLOYMENT STATUS, SEX, AND SEGREGATED FAMILY DECISION NORMS. SURVEY FINDINGS INDICATE THAT WITH RESPECT TO BOTH PHYSICAL PUNISHMENT OF ONE'S CHILDREN AND SPOUSAL VIOLENCE, A SPOUSE'S VIOLENT BEHAVIOR HAS A MUCH GREATER IMPACT ON THE RESPONDENT'S VIOLENCE THAN THE RESPONDENT'S OWN ATTITUDES ABOUT VIOLENCE. WITH RESPECT TO SPOUSAL VIOLENCE ALONE, THE CONSISTENCY BETWEEN ATTITUDE AND BEHAVIOR IS GREATEST AMONG THOSE RESPONDENTS WHO HAVE PROVIOLENCE ATTITUDES AND A VIOLENT PARTNER. CONSISTENCY BETWEEN ATTITUDE AND BEHAVIOR DEPENDS NOT SIMPLY ON A PERSON'S ATTITUDE, BUT ALSO ON SOCIAL STRUCTURE FORCES WHICH REINFORCE OR INHIBIT VIOLENT BEHAVIOR. FOR EXAMPLE, LOW-INCOME HUSBANDS ARE LESS IN A POSITION TO LIVE UP TO THEIR OBLIGATIONS AS PROVIDERS AND HEADS OF FAMILIES THAN ARE MIDDLE-CLASS HUSBANDS. THEIR WIVES, WHO ARE FORCED INTO THE LABOR MARKET, ARE LESS LIKELY TO RECOGNIZE THE MALE AS THE HEAD OF THE HOUSE THAN THEIR MIDDLE-CLASS CONTERPARTS. IN CONTRAST, AMONG HIGH-INCOME HUSBAND, ATTITUDES ABOUT SPOUSAL VIOLENCE SHOWS LITTLE RELATION TO BEHAVIOR. ALTHOUGH THEY MAY APPROVE OF SLAPPING A WIFE UNDER CERTAIN CIRCUMSTANCES, THEY ARE UNIFORMLY LOW IN ACTUAL SLAPPING. IT IS SUGGESTED THAT THIS IS BECAUSE HIGH-INCOME HUSBANDS HAVE ECONOMIC AND PRESTIGE RESOURCES WHICH ALLOW THEM TO CONTROL THEIR WIVES WITHOUT THE NEED TO USE FORCE. REFERENCES ARE INCLUDED IN THE STUDY. (AUTHOR ABSTRACT MODIFIED--LWM).