Standardized household interviews were conducted with 1,286 regular drinkers and 498 nondrinkers who ranged in age from 20 to 60 years. Interview data were obtained from a final sample of 1,374 married, separated, and divorced respondents. Factors in the analysis included social bias, financial stress, sex role traditionalism, and alcohol consumption. Observed ethnic differences on factors likely related to spousal violence were striking. Minority respondents, especially blacks, were more socioeconomically disadvantaged and had significantly more traditional sex role orientations than whites. Whites and blacks were more frequent drinkers than Mexican Americans. Females had less traditional sex role orientations than males and were generally lighter drinkers. Among married respondents, females reported more frequent drinking by spouse than did males, and Mexican Americans reported more frequent drinking by spouse than did whites or blacks. Married black females were more likely to report having been beaten than whites, while Mexican American females were less likely than whites to report spouse beating. Generally, ethnic differences in reports of beating and of being beaten were less pronounced among males than females. The quantity of alcohol consumed rather than frequency or total volume appeared to be the best predictor of spousal violence. While violence was generally greater among high quantity drinkers, particularly females, than among abstainers, violence was actually somewhat greater among lighter drinkers than among high quantity drinkers. Findings raise questions about simplistic socioeconomic status or financial stress explanations of ethnic differences in spousal violence. Further, curvilinear effects of alcohol quantity and spouse drinking on reported spousal violence question simple disinhibition arguments and suggest the need for further research on couple dynamics. 27 references and 3 tables