In the literature, the nature of the relationships between memory processes and summary evaluations is still a debate. According to some theoretical approaches (e.g., two-memory hypothesis; Anderson, 1989) retrospective evaluations are based on the impression formed while attending to the to-be-assessed stimuli (on-line judgment) no functional dependence between information retrieval and judgment is implied. Conversely, several theories entail that judgment must depend, at least in part, on memory processes (e.g., Dougherty, Gettys, & Ogden, 1999; Schwarz, 1998; Tversky & Kahneman, 1973). The present study contributes to this debate by addressing two important issues. First, it shows how more comprehensive memory measures than those used previously (e.g., Hastie & Park, 1986) are necessary in order to detect a relationship between memory and retrospective evaluations. Secondly, it demonstrates how memory strategies influence the relationship between memory and judgment. Participants recalled lists of words, after having assessed each of them for their pleasantness. Results showed a clear association between memory and judgment, which was mediated by the individual strategies participants used to recall the items.