The evaluation included an interview, a medical evaluation, an interview of the primary nonoffending caretaker, and administration of the Trauma Symptom Checklist for Children. At the time of the analyses, external evidence of sexual abuse was available in 30 percent of cases, abnormal medical examinations had been found in 16 percent of cases, and perpetrator confessions had been obtained in 7 percent of cases. Other types of evidence of sexual abuse were present in an additional 6 percent of cases. Disclosures made by the children during their interviews were classified as credible, partially credible or noncredible; their denials of having been abused were classified as credible or noncredible. The findings indicate that children's disclosures and denials can be influenced by variables including sex and race, history of neglect, level of maternal support, and need to deny the abuse or its impact. The authors conclude that self-reported symptomatology is not a reliable a litmus test for child sexual abuse. 2 tables, 1 figure, 5 notes, and 30 references