Many theories have been proposed to explain the blaming the victim phenomena and to predict who is most likely to engage in this behavior. This study hypothesized that since behavioral blame of a victim is self-protective for an observer, this type of blaming will be evidenced more by those who are personally and situationally similar to the victim than by subjects who are situationally similar but personally dissimilar. The latter were expected to utilize characterlogical blame more than the former. These predictions are congruent with the tenets of defensive attribution theory once the important distinction between behavioral and characterlogical blame is taken into account. Subjects (N=30) were female undergraduate students at St. Bonaventure University (New York) who responded to a vignette describing a rape in the local area, with a victim named either John or Sue providing the gender manipulation. Support was not obtained for the reformulation of defensive attribution theory represented by the proposition that subjects of the same gender as victims would attribute more behavioral and less characterlogical blame to the victim than subjects of the opposite gender. The fact that in all conditions subjects preferred behavioral to characterlogical blame offers a possible explanation for the failure to support the hypotheses. The data yielded a significantly greater attribution to chance for scenarios involving male victims rather than female victims. (LLL)