The arrests were made by the Detroit Police Department in 1989. The research combined six victim characteristics to create a scale indicating the stereotypical image of a genuine victim. These factors included the relationship between the victim and the suspect, whether the victim screamed during the attack, whether the victim physically resisted the suspect, whether the victim reported the crime to the police within 1 hour, and whether the police file continued information relating to the victim's moral character. The research also controlled for four evidence factors: witness, injury, physical corroborating evidence, and weapon. The analysis focused on the effects of the genuine-victim scale and the evidence factors on prosecutors' charging decisions. Cases with child victims and cases with adolescent or adult victims were analyzed separately. Results revealed that the genuine-victim scale did not influence charging decisions in cases with child victims. In contrast, this scale was the only significant predictor in cases involving adolescent or adult victims. None of the four evidence factors affected charging in cases with adult victims; only one of these factors was related to charging in cases involving children. Findings suggest that prosecutors attempt to avoid uncertainty by screening out sexual assault cases that are unlikely to result in a conviction due to questions about the victim's character, the victim's behavior, and the victim's credibility. Tables, notes, case citation, and 45 references (Author abstract modified)