Introduction: The present study examined the longitudinal association among drinking problems, drinking to cope, and degree of intimate partner violence (IPV). Two competing models were tested; the first model posited that drinking to cope leads to greater drinking problems and this subsequently leads to more violence in the relationship (an intoxication-violence model). The second model speculated that violence in the relationship leads to drinking to cope, which in turn leads to greater drinking problems (a self-medication model). Methods: Eight hundred and eighteen undergraduate students at a large north-western university participated in the study over a two year period, completing assessments of IPV, alcohol related problems and drinking to cope at five time points over a two year period as part of a larger social norms intervention study. Results: Analyses examined two competing models; analyses indicated that there was support for the self-mediation model, whereby people who have experienced violence have more drinking problems later, and this association is temporally mediated by drinking to cope. Discussion: The current results are discussed in light of past research on the self-medication model.