In computer-assisted surgery, efficacy relies on the overall precision of the method, of which the tracking technology is an integral feature. Does electromagnetic tracking perform clinically as well as standard optical tracking technologies? A pilot study using a computer-assisted surgery system and one lower extremity from an embalmed cadaver evaluated the mechanical axis, the transepicondylar axis, and the anteroposterior axis of Whiteside (anteroposterior axis). Using three-dimensional computed tomography and direct anatomic measurements, the baseline value for the mechanical axis was 4.9 degrees varus and the tibial shaft axis was 4.6 degrees varus. All tests were performed in a standard operating room using an imageless referencing protocol. Repeatability of one surgeon performing eight trials revealed optical mechanical axis mean of 5.8 degrees varus (standard deviation, 0.3 degree) and electromagnetic mechanical axis mean of 5.3 degrees varus (standard deviation, 0.9 degree); reproducibility of three surgeons performing eight trials each revealed optical mechanical axis mean of 6.3 degrees varus (standard deviation, 0.6 degree) and electromagnetic mechanical axis mean of 5.2 degrees varus (standard deviation, 0.8 degree). Precision was satisfactory for both optical and electromagnetic tracking for mechanical axis assessment, but outliers were identified with electromagnetic tracking causing concern for efficacy. Assessment of the transepicondylar or the anteroposterior axis measurements was not satisfactory with either the optical or electromagnetic system.