These children, along with matched controls from the original sample, were re-interviewed 16 months later to determine whether their interactions with the dolls changed over time. Items of interest included the frequency of the childrens' sexualized and avoidant behaviors with the dolls as well as their knowledge of the mechanics of intercourse and possible sources of such knowledge. Findings showed that changes over time occurred in all groups. Changes in the children's behaviors may be explained by cultural, maturational, and socialization factors. Although demonstrations of sexual intercourse by nonreferred, presumably nonabused children with the dolls are relatively uncommon, they do occur and, therefore, cannot be considered definitive markers of abuse; however, such demonstrations do raise the index of suspicion. If a child shows intercourse between the dolls, the interviewer should explore how the child learned that behavior. In order for the behavior to support an abuse history, the interviewer must obtain information that clearly indicates the behavior had happened to the child or that the child was exploited as a passive participant in sexual activities. Likewise, extreme avoidance of the dolls can result from cultural and developmental influences and is not a definitive marker of abuse. Other implications for interviewers are also discussed. 6 tables and 19 references