BACKGROUND: The objective of this study was to assess limb and component alignment after total knee arthroplasty (TKA) on long leg X-rays and to compare the use of iAssist, an accelerometric based computer-assisted device (CAD), with conventional jigs. METHODS: We prospectively recruited 56 consecutive patients undergoing primary unilateral TKA by a single surgeon into this study. In the first 28 patients iAssist navigation system was utilized and in the following 28 conventional jigs were used. The groups were comparable with regard to age, sex distribution, body mass index and preoperative hip-knee-ankle (HKA) angle. Our aim was to restore neutral coronal alignment and a five degree tibial slope. RESULTS: The mean postoperative mechanical axis was 179.4° in the CAD group and 180.1° in the conventional group (P = 0.187). There were five TKAs deviating more than three degrees from neutral (18.5%) in the conventional group and none in the CAD group (P = 0.051). The mean femoral mechanical-condyles-angle was significantly closer (P < 0.001) to our target of 90° in the conventional group but contained more outliers (P = 0.67). The sagittal tibial slope was closer (P = 0.047) to our target of 85° in CAD-navigated TKAs with fewer outliers (P < 0.0001). The Oxford Knee Score showed comparable (P = 0.271) and good clinical outcome in both groups. The mean operation time was significantly longer in the CAD group (P < 0.001). CONCLUSIONS: This study shows that the use of iAssist accelerometric CAD facilitates comparable good leg alignment after TKA in the frontal plane and more accurate tibial slope with fewer outliers in the sagittal plane compared with a conventional technique. Our operation time was longer with CAD.