FOLLOWING EITHER A POSITIVE OR HIGHLY INSULTING EVALUATION OF THEIR PERFORMANCE ON A BLOCK ASSEMBLY TASK, SUBJECTS WERE ASKED TO INDICATE THEIR PREFERENCES AMONG FOUR FILMS--TWO VIOLENT AND TWO TRANQUIL. IN ADDITION, SUBJECTS WERE GIVEN REASON EITHER TO ANTICIPATE A CHANCE TO EVALUATE THEIR EVALUATOR, TO ANTICIPATE A CHANCE TO EVALUATE A THIRD PERSON, OR NOT TO EXPECT TO ACT AS AN EVALUATOR. MEAN PREFERENCES FOR VIOLENT FILMS WERE NOT SIGNIFICANTLY HIGHER IN THE INSULT (ANGER) CONDITON THAN IN THE NO INSULT CONDITION. NOR WAS THERE A SIGNIFICANT DIFFERENCE ACROSS LEVELS OF FILM CONTENT WITHIN THE INSULT CONDITION. HOWEVER, THE FINDINGS DO SUGGEST THAT VIOLENCE IS MORE LIKELY TO BE SELECTED FOR VIEWING (RELATIVE TO NONVIOLENCE) BY AN ANGRY PERSON THAN BY A NONANGRY PERSON. NO SIGNIFICANT EFFECT OF THE AGGRESSION OPPORTUNITY VARIABLE WAS OBSERVED. POSSIBLE EXPLANATIONS FOR THE LACK OF DIFFERENCE BETWEEN VIOLENT AND NONVIOLENT FILM CONTENT PREFERENCE IN THE INSULT CONDITION ARE EXPLORED. SUPPORTING DATA ARE INCLUDED. (LKM)