The children were all outpatient clients assessed by the author over a period of approximately 5 years at a children's mental health center in Windsor, Ontario, Canada. All children were referred to the center mainly in response to their displaying externalized behavior problems such as aggression and noncompliance. The research focused on responses with overt sexual contents, sexual symbols, and near-sexual contents. Results suggested that these dependent variables, in combination, could differentiate sexually abused children from other clinical groups. Other Rorschach indicators, according to the Exner scoring system, were generally not able to differentiate abused from nonabused children. In addition, behavior ratings completed by the parent indicated that for males, higher levels of behavior problems were associated with sexual abuse. Tables and 29 references (Author abstract modified)