BACKGROUND The aim of this study was to assess the long-term performance of a cemented total knee replacement utilising an All Polyethylene Tibial (APT) component and in addition to perform an engineering analysis of any failures to help refine surgical technique. METHODS A total of 26 patients had a total knee replacement performed using a cemented Depuy Press Fit Condylar (PFC) APT component and a cruciate retaining femoral component. At final review all patients were assessed using The Knee Society Score together with radiographs. An engineering analysis simulated loading conditions of the implants that failed and these were compared with the performance of a modular metal-backed Tibial (MBT) component. RESULTS A total of 20 patients were reviewed at mean time of 116 months following surgery. Knee Society Knee Scores and Function Scores in this cohort were 84/100 and 58/100, respectively. Two patients required revision for tibial component failure. Pre-operatively both had valgus deformities and in each case the tibial tray had been lateralised leaving a gap on the medial side where the APT component had no rigid support. The engineering analysis demonstrated that the volume of highly strained cancellous bone was greater in the APT design compared with the MBT design when a model with a 3 mm medial gap was loaded. The stiffer MBT base plate acted more rigidly and shielded the stress applied to the proximal tibial cancellous bone. CONCLUSIONS The APT component demonstrated satisfactory clinical and radiographic performance at long-term follow up. Appropriate cortical support of the APT component is important. The implant should be used with a degree of caution in patients with severe deformities and osteoporosis.