Young children, young adults, and older adults witnessed a scene in groups of four same-aged acquaintances. Unbeknownst to the group, a technology allowed the scene to appear differently to one member (minority) than to the others (majority), which obviated the need to rehearse confederates or to artificially provide misinformation for the report conformity effect. After viewing, participants had public recollections, and 3 days later, their cued memory and confidence (for adults) were tested privately. Majority members' reports influenced the minority members' but only for adults, not for children (17% conformity compared with 35% for adults), thus providing evidence of developmental reversals in memory reports of verbatim details. Answer changes between the sessions were dramatically higher for minority participants at all three ages (6.7 vs. 66.7% for children, 10 vs. 50% for younger adults, and 26.7 vs. 63.3% for older adults). We discuss the implications of these findings for questioning cowitnesses.