The present experiment was designed to test a metamemory interpretation of the self-choice effect, or memorial advantage for verbal materials that are chosen by the subjects rather than by the other subjects or the experimenter. Third grade and college subjects were presented with two lists of words, some of which they were instructed to remember and some of which they were instructed to forget. Half of the to-be-remembered (TBR) words were chosen by the subjects, and the remainder were selected randomly by the experimenter. During the test phase, subjects were asked to recall and recognize the all of the TBR words. For recall, the self-choice effect emerged for college students, but not for third graders. For recognition, the self-choice effect was obtained for both college students and third graders. These results are consistent with an explanation of the self-choice effect’s being due to subjects' use of metamemory.