Amnesia, as a central descriptive and diagnostic feature of dissociative identity disorder (DID), has received little empirical study. The few published studies are generally consistent in finding that direct tests of memory (e.g., recall and recognition) produce reports of interidentity amnesia but less transparent indirect tests of memory tend to show evidence of memory transfer between identities. Such findings highlight the need for more objective measures of memory in DID and raise questions concerning the nature of amnesia in DID. At present, empirical research fails to unequivocally substantiate patients' claims of amnesia between identities, and reports of such amnesia should not be regarded as conclusive in legal proceedings. The authors propose that psychophysiological measures of memory may provide such an objective measure and can further illuminate the nature of the reported memory deficits in DID.