This study investigated the role of metacognition in time-based prospective memory (PM). The main purpose of the study was to explore if metacognition could explain the Age PM Paradox. 57 young adults (M = 22) and 54 older adults (M = 68) performed a laboratory and a naturalistic PM task. As a measure of metacognition, they also gave item-by-item judgments- of-learning (JOLs) for both the prospective and retrospective components of PM. Results confirmed the Age PM Paradox and showed a different age effect for the retrospective and prospective components. Regarding JOLs, older adults gave better predictions than young adults. The Age PM paradox is partially explained by individuals' metacognitive awareness: there is a link between metacognition and performances in the field but not in the lab. Moreover, we found a mediation effect in which participants who overestimated their PM performance obtained poorer PM performance because they controlled less the time in the lab (for old adults) or the time schedules in the field (for young adults). The link between metacognition and PM performance is discussed.