Previous research has reported that young adults are better at eyewitness face recognition than are older adults. However, these studies have used young adults as culprits and fillers. We explore how the relative ages of the witness and the culprit influence eyewitness accuracy in 2 experiments. In the first experiment, young (18-25 years old) and older (35-55 years old) adults each saw 4 crime videos. In 2 the culprit was a young adult and in 2 the culprit was an older adult. Participants were more accurate at identifying the culprit when viewing culprit present lineups comprising people of their own age: an "own age bias" analogous to the own race bias. In the 2nd experiment, using a similar procedure, young (18-33 years old) and older (40-55 years old) adults viewed both culprit present and culprit absent lineups. The results of the first experiment were replicated for the culprit present lineups. However, no own age bias was found for the culprit absent lineups. Implications for police procedures dealing with cross-generation identifications are discussed.