The authors examined 98 total knee arthroplasties in 73 patients who were 80 years of age or older at the time of surgery (average, 82 years; range, 80–90 years). The follow-up period averaged 4.5 years (range, 2–12 years). The patients were divided into two groups based on their tibial component design. There were 38 all-polyethylene tibial components in 28 patients and 60 metal-backed tibial components in 45 patients. There were 61 (62%) excellent, 30 (31%) good, 2 (2%) fair, and 5 (5%) poor results. Three of the five poor results required revision for septic failure. Of the knees with an all-polyethylene tibial component, 20 (53%) were rated as excellent, 15 (39%) as good, and 3 (8%) as poor. One of these knees rated as poor required revision for septic failure. The knees with a metal-backed tibial tray had 41 (68%) rated as excellent, 15 (25%) as good, 2 (3%) as fair, and 2 (3%) as poor. Both of the knees with poor results required revision for septic failure. Stratifying the results by component composition revealed 97% survival for both types of tibial trays. These results were obtained at 12 years for the all-polyethylene components and at 8 years for the metal-backed prosthesis. In conclusion, the authors believe that total knee arthroplasty is a reliable and durable procedure in the treatment of knee arthritis in the elderly. Elderly patients may represent a special case because they are generally less active than younger patients and may place less stress on their prosthesis. This may place them at less risk for mechanical failure and may make an all-polyethylene tibial component a viable option in the thin, elderly patient, who is usually less active.