Successful priming of amnesics’ performance on a wide variety of implicit mem­ ory tasks has been a source of fascination for investigators of memory for nearly a decade (Graf & Schacter, 1985; Graf, Squire, & Mandler, 1984; Jacoby, Toth, Lindsay, & Debner, 1992; Roediger, 1990; Schacter, 1987; Squire, 1987). These researchers have incorporated the finding that amnesic patients perform as well as normal controls on implicit tasks into their theories, relying on the phe­ nomenon as independent neurological evidence for the existence of the dichoto­ my of memory systems. This interest has produced an outpouring of research on memory disorders unparalleled within other decades of psychological research. Despite this abundance of interest, data, and methodology, the payoff for a further understanding of amnesia itself, from an information processing point of view, has been disproportionately small. This is largely because, by the very nature of the investigation, what is being studied is not amnesia. Instead it entails determining what it is that amnesics can do without directing any further atten­ tion toward exploring the underlying processes that contribute to those memory tasks they cannot do. Being discovered as a viable avenue of proof has entailed a certain amount of patience on the part of memory disorders researchers who are eager to get back to studies of deficits. It is probably similar to the situation that occurs when the public discovers a particular artist, who then finds it difficult to return to her studio because of the parties she is obligated to attend. But now it is time to return to the exploration of amnesics’ memory disorder, and it is the purpose of this chapter to suggest a way back that is not necessarily an abrupt change but rather a transition from the pure study of implicit memory to an investigation of processing disturbances experienced by amnesics even during the course of their apparently normal implicit performance.