Objective: The present study investigated 27 sexually abused children's reports about abuse given in the context of police interviews. All abuse cases had been verified (with, e.g., photographs or video films), proving that abuse had occurred. Method: The interviews with the children were analyzed regarding amount and type of information reported, and the frequency of denial and avoidance. Furthermore, children's reporting on different interview occasions was investigated. Results: Children reported significantly more neutral information from the abusive acts "per se" than sexual information. The children were also highly avoidant and, on several occasions, denied that (documented) sexual acts had occurred. Furthermore, the second and third interviews generated twice as many (new) sexual details as the first interview. The children also produced more denials and avoidances at the first interview compared to subsequent interviews. Conclusions: The present study indicates that sexually abused children may be highly resistant to reporting about the abuse in police interviews, and that two or three interviews may be needed to enable children to give complete and informative reports. It is of vital importance that professionals within the legal system be aware of this problem when conducting child interviews and when evaluating the reliability of child sexual abuse reports. (Contains 5 tables.)