100 patients undergoing total knee arthroplasty were randomized to tibial resection using either an extramedullary device or a portable, accelerometer-based navigation device. The purpose was to evaluate accuracy of tibial component alignment. Results revealed that the use of a navigation device offered better tibial component alignment and differences between intraoperative goals and postoperative alignment were smaller compared to an extramedullary device. Duration of tibial resection, however, was significantly longer with a navigation device. Femoral component alignment was similar with both devices.