Six- to nine-year-old children participated in a training experiment concerning study and recall of categorized items. Explicit feedback given during training sessions was devoted to emphasize the relationship between strategy use and recall performance. Verbal descriptions of the contents of these training sessions were asked before a delayed post-test, given two weeks later. It appears from the results that: (1) The degree of understanding of the information provided by training with explicit feedback, as indicated by subjects’ verbal descriptions, is a good predictor of strategy maintenance on the post-test. Explicit feedback provides subjects with information about the relationship between procedures and results. However, only those subjects who were able to re-elaborate this information were able to construct more permanent metacognitive knowledge; (2) Such a form of cognitive re-elaboration appears to be necessary for acquired metacognitive knowledge to have a regulatory effect on subsequent behavior; and (3) The level of a subject’s prior metacognitive knowledge seems to be one of the determinants in the ability to efficiently integrate new metacognitive knowledge by re-elaboration of metacognitive experiences.