One hundred consecutive cases treated with ankle arthroplasty for osteoarthritis (OA) or rheumatoid arthritis were followed prospectively and annually for up to 15 years. Survivorship analysis was performed, with the endpoint being prosthesis revision or change to arthrodesis. Patients who were younger than 50 years at the first implantation constituted one group (group A, 30 ankles). The other group (group B, 70 ankles) consisted of patients aged 50 years or older at the first implantation. All patients were assessed clinically according to the Kofoed Ankle Score. The distribution of OA/rheumatoid arthritis in group A was 18/12, and in group B it was 43/27 (not significant). The median age in group A was 46 years (range, 22-49 years), and in group B it was 63 years (range, 51-83 years). In group A, one case was revised, and three cases were converted to arthrodesis after a median of 5 years (range, 5-9 years). In group B, four cases were revised, and four cases were converted to arthrodesis after a median of 5.5 years (range, 2-8 years). The results of cases with traumatic OA did not differ between groups A and B. It was concluded that the results of ankle arthroplasty were of equal quality in patients younger than 50 years and those who were older.