People differ in how quickly they forget, yet whether beliefs about one’s own memory relate to the rate of forgetting has not been directly tested. This study examined whether pre-learning concurrent memory self-efficacy predicts individual differences in forgetting rate over one week, beyond general memory self-efficacy. Seventy-five adults completed a general memory self-efficacy measure and predicted their recall of a 15-word list, with a confidence rating, before and again after learning it across five trials. Recall was tested after a short delay and again at an unannounced telephone follow-up about one week later, from which a proportional forgetting rate was computed. Neither general nor pre-learning concurrent memory self-efficacy was related to forgetting rate, and confidence added nothing beyond the numerical predictions. In a post hoc comparison, only the prediction made after learning was related to forgetting rate. These findings are consistent with accounts that place the influence of memory self-efficacy at learning rather than in the retention interval and suggest that judgments of one’s own retention may become informative only after experience with the material.