The discussion focuses on abuse memories recovered through "memory work" with people who previously did not believe they had been sexually abused as children. The authors first discuss the cultural context in which the recent popularity of memory work developed. They then review scientific evidence of the validity of recovered histories of childhood sexual abuse. Central considerations include the base rates of child sexual abuse and of forgetting such abuse, practitioners' ability to discriminate between clients with and without hidden histories of childhood sexual abuse, the susceptibility of memory to suggestion, and factors that increase the difficulty of discriminating between veridical and illusory memories. The authors argue that memory work can yield both veridical memories and illusory memories or false beliefs. In their view, allegations of childhood sexual abuse based solely on recovered memories should be viewed with skepticism if there is evidence of exposure to suggestive information or memory recovery techniques. The likelihood of claims based on recovered memories being valid can only be crudely estimated by weighing evidence regarding how memories were recovered, the nature and clarity of the recovered memories, the plausibility of the alleged events being forgotten, the plausibility of recovering the memories, and the base rate of the alleged type of abuse. The paper concludes with a consideration of the implications of the best available evidence regarding recovered memories for professional psychology, public policy, and the law. 316 references