The occurrence of an own-age bias on age estimation of faces was assessed by comparing children (10-14), young (20-30) and older (65-75) adults performances. Participants from each age group estimated both in-group and out-group faces. Overall, the age of childrens faces was more accurately estimated than the age of both young and older adult faces. Moreover, the age of young adult faces was more accurately estimated than that of older adults. A significant Age of participants X Age of face stimuli interaction was revealed. Although we did not observe a crossed interaction where each age group would have been more accurate for in-group estimation than for out-group estimation, present results reveal the occurrence of an own-age bias in age estimation. There was no significant difference between the three groups when estimating child faces, but young and older adults were more accurate for in-group estimation than the other groups of participants.