The dominant theme of the reviewed research is that alcohol consumption has a causal link to violent behavior. Numerous theories are presented to explain this causal link. Most come under the following theoretical models: an expectancy model, a physiological disinhibition model, or an indirect causal model. The expectancy model reasons that alcohol-induced aggression stems from the learned beliefs about the behaviors that can be expected following alcohol consumption, rather than the pharmacological effects of ethanol itself. Thus, the violent drinker simply expects to behave aggressively after drinking; however, the author concludes that the findings presented in this article suggest expectancy has a minimal effect on aggression after alcohol consumption. This leads to the conclusion that alcohol's pharmacological properties explain aggressive behavior while intoxicated. Whereas, the physiological disinhibition model argues that alcohol inevitably has the effect of diminishing inhibitions against an innate human tendency toward aggression, the indirect causal model argues that the properties of alcohol constitute only one factor that indirectly contributes to aggressive behavior under its influence. The influence of alcohol in reducing customary behavioral controls can combine with external factors such as provocation, frustration, and social conflicts to stimulate aggression that would be controlled or reduced if the effects of alcohol were not present. The author argues that this latter theoretical model best fits research findings. Future research directions are suggested. 1 figure, 6 tables, 4 notes, and 51 references