The aim of this study was to develop and test a robust approach to apply a joint coordinate system (JCS) to imaging data sets of the glenohumeral joint and to reconstruct the kinematics with six degrees of freedom (6DOF) in order to investigate shoulder pathologies related to instability. Visible human data were used to reconstruct bony morphology. Landmarks were used to define axes for body-fixed Cartesian coordinate frames on the humerus and scapula. These were applied to a three-cylinder open-chain JCS upon which the humeral 6DOF motions relative to the scapula were implemented. Software was written that applies 6DOF input variables to rotate and translate the nodes of the surface geometry of the humerus relative to the scapula in a global coordinate frame. The instantaneous relative position and orientation of the humerus for a given set of variables were thus reconstructed on the bone models for graphical display. This tool can be used for graphical animation of shoulder kinematics, demonstrating clinical assessments, and allowing further analysis of the function of tissues within the joint.