In the past ten years there has been increased use of six-degree-of-freedom instrumented spatial linkages for the measurement of biological joint motions. In spite of the increased popularity, little information has been reported on the accuracy of these devices. In this paper, we present a two-part investigation of the accuracy of the instrumented spatial linkage when used to measure knee joint kinematics. In the first part, we present the results of a theoretical analysis and an experimental determination of the errors associated with spatial linkage systems. In the second part, we describe the errors associated with a bi-planar X-ray system used to obtain the coordinate transformation between the linkage ends and coordinate systems located in the bones comprising the joint. We found that the theoretical error analysis consistently overestimated the actual measurement error, thus providing an unreliable estimate of errors. The experimental study of both the linkage system and the bi-planar X-ray system demonstrated that the accuracy of displacement measurement is insensitive to large systematic errors in position measurement.