Abstract Introduction Age estimation performances may be influenced by group biases. Objective This study investigated whether we are more accurate at estimating the age of people from one's own-age than the age of younger or older people. Method Children, young and older adults’ performances at estimating both in-group and out-group faces were compared. Results A significant “Age of participants” × “Age of face stimuli” interaction was revealed. Moreover, the age of children's faces was more accurately estimated than the age of young and older adults’ faces by the three groups of participants. Conclusion The present results revealed the occurrence of an own-age bias for children, young and older adults in age estimation. Several explanations to this own-age effect are discussed.