Current rehabilitation protocols and surgical procedures of the elbow are not scientifically justified. As a result. the outcomes of many upper extremity disorders remain suboptimal. The purpose of this study was to develop a clinically relevant manner to study upper extremity kinematics and to determine the optimal co-ordinate system to quantify this motion. An unconstrained loading apparatus, incorporating a loading frame, pneumatic actuators, a tendon alignment unit and an electromagnetic tracking system, was developed. In testing of ten cadaveric specimens, active motions were readily simulated using computer-controlled sequential loading. In all cases. simulated active motion was more repeatable than passive motion for supinated flexion and forearm supination (p<0.05). This suggests that studies which use simulated active loading to initiate joint motion provide more reproducible, clinically relevant kinematics than passive motion. Upon investigating different kinematic co-ordinate systems, it was detemined the Euler Angle Theorem, with all bone co-ordinate systems detemined from digitization points and independent of each other, provides the optimal description of elbow kinematics. Incorporation of simulated active loading will be a valuable laboratory tool to allow refinements in surgical procedures and rehabilitation protocols. These results will allow further investigation of both stable and unstable elbow kinematics and permit the results to be reported in both a mathematical and clinically acceptable manner.