Measurement of upper limb motion is problematic, not least because of the large range of path dependent description of motion of the joints, and the multiple non-cyclical unstandardised motion tasks measured. Furthermore, appreciation of shoulder motion specifically is obscured by overlying soft tissue. In order to satisfy the complexity of a clinically useful model of the movement of the joint, input data must be acquired from a set of pre-determined movements using a non-invasive technique with a high level of accuracy. Descriptive and predictive modeling of the glenohumeral joint requires input of high-fidelity data into a 6 degree of freedom representation, without which, the application of the tool is of limited clinical significance to the advancement of both operative and non-operative management of shoulder pathology. Electromagnetic, linkage and radiographic techniques have previously been used, however, an optimal solution is yet to be described.