Objective: This study provided the first direct test of the cognitive underpinnings of the attention-allocation model and attempted to replicate and extend past behavioral findings for this model as an explanation for alcohol-related aggression. Method: A diverse community sample (55% African American) of men (N = 159) between 21 and 35 years of age (M = 25.80) were randomly assigned to 1 of 2 beverage conditions (i.e., alcohol, no-alcohol control) and 1 of 2 distraction conditions (i.e., distraction, no-distraction). Following beverage consumption, participants were provoked via reception of electric shocks and a verbal insult from a fictitious male opponent. Participants' attention allocation to aggression words (i.e., aggression bias) and physical aggression were measured using a dot probe task and a shock-based aggression task, respectively. Results: Intoxicated men whose attention was distracted displayed significantly lower levels of aggression bias and enacted significantly less physical aggression than intoxicated men whose attention was not distracted. However, aggression bias did not account for the lower levels of alcohol-related aggression in the distraction, relative to the no-distraction, condition. Conclusions: These results replicated and extended past evidence that cognitive distraction is associated with lower levels of alcohol-related aggression in highly provoked males and provide the first known cognitive data to support the attentional processes posited by the attention-allocation model. Discussion focused on how these data inform intervention programming for alcohol-related aggression. (Contains 2 footnotes and 2 figures.)