From 1974 until the end of 1988, 204 primary Mayo total ankle arthroplasties were performed at the Mayo Clinic. By means of actuarial analysis, we determined the cumulative rates of survival with failure (defined as removal of the implant) as the end point. The average duration of follow-up was nine years (range, two to seventeen years). By applying the Cox proportional-hazards general linear model, we identified two independent variables that were associated with a significantly higher risk of failure: a previous operative procedure on the ipsilateral foot or ankle and an age of fifty-seven years or less. The overall cumulative rate of survival at five, ten, and fifteen years was 79, 65, and 61 per cent, respectively. The probability of an implant being in situ at ten years was 42 per cent for patients who were fifty-seven years old or less and who had had previous operative treatment of the ipsilateral ankle or foot and 73 per cent for those who were more than fifty-seven years old and who had had no such previous operative treatment. We do not recommend the use of the Mayo total ankle arthroplasty, particularly in younger patients who have had a previous operative procedure on the ipsilateral ankle or foot.