People cannot remember what happened when they were next-in-line to perform. Theorists have wondered whether this memory deficit reflects a failure to encode or an inability to retrieve preperformance events. These accounts are contrasted in a pair of studies. In Experiment 1, preperformance memories are assessed with free and cued recall. Although our semantic cues strongly facilitate access to memories, they do not moderate the next-in-line effect. In Experiment 2, subjects are told-either before or after performing-to make a special effort to remember preperformance events. If instructed afterward, subjects display the usual memory deficit. If instructed beforehand, they reverse the deficit and show a superior preperformance recall. The next-in-line effect is a failure at encoding, not at retrieval.Students who expect to be called on may not remember their teachers' remarks. In a study of this phenomenon, Brenner (1973) had a group of subjects take turns reading words aloud. On a subsequent test of free recall, subjects could not remember words read immediately before their reading performance. Relative to subjects who merely listened to all the words, readers displayed a next-in-line effect, a deficit in recalling events up to 9 s prior to performance.Researchers have described characteristics of the next-in-line effect. Brenner (1973) showed that preperformance memory deficits depend on performance difficulty, with more difficult performances yielding larger deficits. Walker and Orr (1976) found that the magnitude of the next-in-line effect is independent of a performer's fear of negative evaluation. Brown and Oxman (1978) suggested that preperformance memory deficits might be a retroactive consequence of performing, much like retrograde amnesia (Detterman, 1975). Bond and Kirkpatrick (1982) discredited this suggestion, finding that subjects who anticipate performance have a memory deficit even if they never perform.In this article, I consider information-processing explanations for the next-in-line effect.I am grateful to Rich Morrison and Nancy Wells for their help with this research.Requests for reprints should be sent to Charles F.