OBJECTIVE Proper alignment of the prosthesis is critical in total knee replacement (TKR) to minimize long-term wear, risk of osteolysis, and loosening of the prosthesis. This study examined the accuracy of lower limb alignment obtained using a kinematic navigation system for TKR, and the extra time needed to adopt this system. METHODS From August 2002 to April 2003, 71 patients with knee osteoarthritis underwent 79 primary TKR operations by the same surgical team. Fifty of these operations were performed with the aid of the CT-free kinematic navigation system, and the remaining 29 were performed with conventional manual methods. Results, including operation time, radiographic alignment of the prosthesis and complications, for the two groups were compared. RESULTS Patients in the kinematic navigation group achieved better accuracy in the coronal plane than the conventional group in terms of postoperative mechanical axis (1.89 +/- 0.63 degrees vs. 3.38 +/- 1.07 degrees ). Less variation was noted in the navigation group (femur: SD 1.88 degrees vs. 7.12 degrees ; tibia: SD 1.54 degrees vs. 2.99 degrees ), although the difference in the mean values was not significant (p = 0.475 and 0.55, respectively). The operation time (from skin to skin) in the navigation group (100.6 +/- 4.3 minutes) was longer than that in the conventional group (92.7 +/- 5.1 minutes; p = 0.027). Two perioperative fractures occurred in the navigation group, both of which were attributed to patient factors as opposed to operation procedures. No major complications such as infection or pulmonary embolism occurred during this study. CONCLUSIONS Use of a kinematic navigation system in TKR provides better accuracy than conventional manual methods. The technique is easy to use, has a short learning curve, and requires an additional operation time of less than 10 minutes. Precise alignment can be achieved with the aid of navigation in most cases.