Repeated and prolonged searches of memory can lead to an increase in how much is recalled, but they can also lead to memory errors. These 3 experiments addressed the costs and benefits of repeated and prolonged memory tests for both young and older adults. Participants saw and imagined pictures of objects, some of which were physically or conceptually similar, and then took a series of repeated or prolonged recall tests. Both young and older adults recalled more on later tests than on earlier ones, though the increase was less marked for older adults. In addition, despite recalling less than did young adults, older adults made more similarity-based source misattributions (i.e., claiming an imagined item was seen if it was physically or conceptually similar to a seen item). Similar patterns of fewer benefits and more costs for older adults were seen on both free and forced recall tests and on timed and self-paced tests. Findings are interpreted in terms of age-related differences in binding processes.