Forty-eight male subjects participated in an aggression task after having received either alcoholic or placebo beverages and after being told that they had received either a high dose or low dose of alcohol. All subjects in the alcohol group actually received the same dose. Measures of aggression were the intensity and duration of shock given to a bogus partner in a reaction time-pain perception task and a score on a posttask attitude questionnaire. Blood alcohol readings were taken three times during the session, and a locus of control scale was administered before and after drinking. For subjects in the alcohol conditions, blood alcohol levels averaged .079% before the task and .085% upon completion of the task. Alcohol resulted in higher aggression scores only on the questionnaire measure, and an interaction between drug and attribution was found for the measure of shock duration. Attribution alone resulted in greater aggression on the measure of shock intensity. Subjects in groups where the attribution was discrepant with the drug condition were the most aggressive on the measures of both shock intensity and duration. These subjects became more external, reflecting a high correlation between their aggression and a stimulus tone. The results of this study suggest that a person's increased aggressiveness when intoxicated results from an interaction situation between alcohol and an alteration in responsivity to provoking stimuli.