In the present study, a single -trial learn-recall design was employed to eliminate the possibility of interference and to provide a clean test of the trace decay theory, which predicts forgetting under impoverished feedback and interpolated noninterfering activity because of covert-rehersal restrictions. (Author)