Two experiments were conducted to determine if eyewitness accuracy was related to the relative ages of the witness and the culprit. In the first experiment, participants were shown four videos of crimes being committed. Participants were then asked to identify the culprits in the videos. The results show that when young culprits and fillers were used, the young adults outperformed the older adults. When the culprits and fillers were older adults, the younger participants did not do better than the older adults. The older adults performed slightly better than the young adults. The second experiment was designed to replicate the own age bias in culprit present lineups and to test whether it occurred also in culprit absent lineups. The same four crime videos were used. Results show that when young culprits and fillers were used in culprit present lineups, younger participants were more accurate at lineup identifications than older participants. When culprits and fillers were older, younger participants were not more accurate than older participants. Older participants were marginally more accurate with culprits of approximately their own age. This replication of the results from the first experiment supplies strong evidence for the existence of an own age bias. An important result of this experiment relates to the culprit absent condition. It was observed that the own age bias only occurred with culprit present lineups, not with culprit absent lineups. Therefore, it appears that older witnesses will not be more likely than younger participants to identify an innocent young suspect, but they will be more likely to fail to identify a guilty young culprit. There are large and justified concerns about the accuracy of lineup identifications. Further replications are necessary to determine the robustness and size of the effect of age bias. 1 figure, 3 tables, 5 footnotes, 43 references