The increasingly frequent clinical reports of the recovery of repressed memories of childhood sexual abuse have drawn both skepticism and support in the analytic community. Two contrasting paradigms are offered to account for the processes by which fully repressed memories are recovered. On the one hand, the analyst's belief that one can reconstruct early traumatic experience creates fertile ground for overt and covert suggestion, which, in conjunction with an anxious patient seeking affiliation, may lead to the production of false memories. On the other hand, the analyst's belief in the likelihood of repressed abuse and that it can be reconstructed may constitute a necessary precondition for the emergence of valid memories. With these beliefs providing an essential holding environment, the recovery of repressed memories of sexual trauma may be an instance of the retrieval of state-dependent memory. Although the preponderance of evidence favors the suggestion hypothesis, the evidence is not conclusive. Which paradigm the analyst adopts, however, carries enormous clinical implications.