The subjects were 72 5- and 7-year-old girls who answered (1) free-recall questions about their visit to the doctor, (2) questions using an anatomically correct doll, and (3) direct and misleading questions about the doctor's appearance and events that occurred during the visit. Analyses of the completeness and accuracy of the children's answers showed that most vaginal and anal touch went unreported in the free-recall section of the interview, but were revealed when children were asked direct doll-aided questions. In free recall and demonstration, an age advantage was shown only for children in the nongenital condition. In this study, false reports of vaginal and anal touch from children in the nongenital condition were nonexistent in free recall and demonstration with dolls; were rare in response to direct questions. In terms of responses to direct questions, there were significant differences between the two age groups one week, but not one month, after the examination. 5 tables, 5 notes, and 36 references