This paper tests whether alcohol consumption is a determi- nant of domestic violence among partners, considering the pos- sible endogeneity of alcohol consumption. Results indicate that drinking increases the probability of domestic violence for inci- dents considered of low level of violence. In contrast, drinking is endogenous for incidents of high and life-threatening levels of violence. These results imply that a tax on alcohol can only guarantee a reduction of the low levels of violence, while for more serious levels of violence the e§ect is uncertain.