People sometimes exhibit a ‘forgot-it-all-along bias’ in which they claim that they have gone for months or years without thinking about certain childhood experiences despite recently recalling those memories. The present study examines memory for memories of childhood experiences, expanding on prior work by using manipulations that require greater reflection when thinking about remembered experiences and when making retrospective metamemory judgments. Age-related differences in memory-for-memory accuracy were also examined. Young (18–20) and older adults (63–89) recalled various events while focusing on emotional or perceptual details for some, and several weeks later were asked to indicate the last time they had remembered various events. Results showed that young adults were more accurate than older adults overall, though both age groups still exhibited a forgot-it-all-along bias that was reduced but not eliminated when a contextual reminder was provided.