This study reports the minimum 5-year follow-up of our experience with the Duracon Total Knee Arthroplasty System. A total of 121 consecutive total knee replacements using the Duracon system (Howmedica, Rutherford, NJ) were performed in 104 patients. Three patients died before the 5-year follow-up and were excluded from the final evaluation. The remaining 118 knees (101 patients) were assessed at a mean follow-up of 65 months (range, 60-80 months). The knee diagnoses were osteoarthritis in 97 patients, rheumatoid arthritis in 2 patients, osteonecrosis in 1 patient, and pigmented villonodular synovitis in 1 patient. The mean age was 70 years (range, 28-85 years). There were no reoperations for aseptic loosening, and there have been no reoperations for patellofemoral problems. At final follow-up evaluation, 112 knees (96%) had good or excellent results, and 6 knees (4%) had poor clinical results or went on to revision. For the surviving knees, the preoperative Knee Society objective score improved from a mean of 52 points (range, 20-72 points) to a final follow-up mean of 94 points (range, 66-100 points). Five knees needed reoperations: 2 knees in 1 patient because of acute hematogenous infection at 12 months, 1 knee because of a supracondylar femur fracture, 1 because of a patellar tendon rupture, and 1 to increase polyethylene thickness because of instability. The lack of aseptic loosening at the minimum 5-year follow-up compares favorably with any cemented or cementless series of knee replacement. The almost complete absence of patellofemoral complications in this series also indicates that the design changes, with particular attention to the trochlea design and patellofemoral contact throughout full flexion, have achieved their intended purpose. The results are encouraging at midterm, awaiting true long-term (15-20 years) follow-up.