Evaluation of child sexual abuse often necessitates interviewing children about genital touch, yet little scientific research exists on how best to obtain children's reports of genital contact. To examine this issue, 72 five- and seven-year-old girls experienced a standardized medical checkup. For half of the children, the checkup included a vaginal and anal examination (genital condition); for the other half, the checkup included a scoliosis examination instead (nongenital condition). The children's memories were later solicited through free recall, anatomically detailed doll demonstration, and direct and misleading questions. The majority of children in the genital condition revealed vaginal and anal contact only when asked directly about it. Children in the nongenital condition never falsely reported genital touch in free recall or doll demonstration; when asked directly, the false report rate was low. Significant age differences in free recall and doll demonstration, found only in the nongenital condition, implicated socioemotional factors as suppressing the reports of older children who experienced genital contact.