Over 5 years (1985-90), 2,384 children were referred for suspected sexual abuse and were included in a prospective database that encompassed demographic information, clinical history, details of disclosure, relationship of perpetrators, the nature and duration of abuse, and medical findings. All children were examined within 7 days of the referral if the abuse occurred over 2 weeks ago, within 48 hours if the abuse occurred within the past 2 weeks, and immediately if the abuse occurred within the last 72 hours. Many of the positive medical findings showed evidence of acute assault that would heal quickly and completely if the examination were delayed. Of the children referred, 69.2 percent were referred after they disclosed abuse, and 30.8 percent had not disclosed abuse but were referred because of behavioral changes, exposure to an abusive environment, or anatomical variations or medical conditions. Of the total sample, 96.3 percent had a normal examination. In the disclosing group 95.6 percent had a normal examination compared with 97.8 percent in the non-disclosing group. The notable result of this study was that such a small percentage of the children had genital evidence of prior trauma from sexual abuse. Decades of research into the medical diagnosis of child sexual abuse, however, indicate that most children remain free of any medical findings that indicate penetrating trauma. The history from the child remains the most important part of any assessment. The medical examination should then proceed with attention to the child's needs and with the anticipation that most examinations will be normal. The medical examination should be interpreted in the context of how children are abused, the child's perception of the abuse, and the process of disclosure. 4 tables and 36 references