This study examined the effects of perceived similarity, behavioral confirmation of sex role stereotypes, and victim gender on attributions of blame and derogation of the victim. College students (N=120) participated in a 2x2x2 factorial design manipulating victim gender, victim behavior (active/passive), and personal similarity (young student/older clerk). Subjects read a vignette describing a police report of an assault and completed questionnaires measuring subjects' perceptions of the crime, attributions of blame, attributions to the victim, and attributions to the attacker. The results revealed that attacks of high similarity victims were perceived as less frequent than those of low similarity victims. Active males and passive females were perceived as having more frerquent attacks than passive males or active females. The victimization of similar females and dissimilar males was perceived as being more serious than the victimization of dissimilar females or similar males. Subjects assigned more blame to the attacker when the victim was passive rather than active. Dissimilar victims, female victims, and active victims were more likely to be blamed for their attacks. Overall, no one attributional model was clearly supported. (NB)