There are large individual differences in the degree of association between the accuracy of memories and subjective confidence in those memories. Are these differences stable within the same test, and between alternate forms of a test? In Experiment 1, college students were tested on 3 recognition memory tasks, then retested 2 weeks later on alternate forms of the same tasks. The relationship between confidence judgments and recognition performance displayed low split-half stability and low alternate-forms stability.Asecond experiment with elderly adults replicated these findings. In a third experiment, college students recalled answers to general knowledge questions and rated confidence in the correctness of each answer. Individual differences in the association between confidence and recall performance were not stable across the odd-and even-numbered items on the test. These data indicate the need for the development of procedures that will produce stable estimates of individuals' metacognitive accuracy.The present research investigated the association between the accuracy ofmemories and a person's confidence in those memories, a topic that has long been of interest in psychology (see, e.g., Seward, 1928;Strong, 1912Strong, , 1913Trow, 1923). The association between confidence judgments and memory performance has been important in several areas ofresearch, and currently is one aspect ofmetacognition being studied along with others such as feelingof-knowing judgments, judgments oflearning, and calibration of comprehension. Also, in the eyewitness memory field, the usefulness of eyewitness confidence as a predictor ofrecall and recognition accuracy has been the subject ofmuch research (e.g., Brigham, 1990;Deffenbacher, 1980;Leippe, 1980;Luus & Wells, 1994;Perfect, Watson, & Wagstaff, 1993;Wells & Murray, 1984).Individuals can vary greatly in metacognitive accuracy, that is, how accurately their metacognitive judgments predict performance on a criterion task. For example, Deffenbacher, Leu, and Brown (1981) reported that individual subjects' point-biserial correlations between confidence judgments and performance on a face recognition task ranged from -.05 to .60. In the research presented in this article, Goodman-Kruskal gamma correlations between confidence and memory performance covered almost the entire possible range, from -1 to +1.The magnitude ofthese individual differences is not confined to studies in which Portions of these data were presented at the annual meeting of the American Psychological Association, Boston, August 1990. We thank R. Crowder, R. Fisher, A. Glenberg, C. MacLeod, T.Nelson, and R. Proctor for their helpful reviews of earlier versions of this article. We also thank 1. Allen, R. Bonner, and D. DeMare for assisting with data collection. Correspondence should be addressed to W. B. Thompson, Department of Psychology, Niagara University, Niagara Falls, NY 14109-2208 (e-mail: wbt@niagara.edu). subjects make postrecall and postrecognition confidence judgments; correlations between feeling-of...