The shoulder complex is prone to numerous pathologies and instabilities due to its large range of motion. The extent of injury is assessed through a series of observations and physical examinations. It is hypothesized that objective kinematic analysis of the shoulder could yield useful functional insights to aid clinical practice. Non-invasive motion analysis techniques to monitor shoulder function have been developed using passive markers; however, accurate measurement of scapula kinematics is problematic because of overlying tissue. The scapula locator is the accepted standard by which alternative non-invasive techniques of scapula tracking are validated. In this study, the viability of using skin-mounted markers to measure dynamic scapula movement is determined. Complete kinematic descriptions of ten healthy shoulders were obtained. Elevations of the glenohumeral joint were similar with both techniques, indicating that the skin marker method is suitable for gathering functional glenohumeral data. The main differences of note are seen at the scapulothoracic articulation where the skin marker method underestimated lateral rotation by more than 50 degrees at maximum elevation. However, the correlation between the two approaches is greater than 0.7, suggesting that it may be possible to derive linear regression models to predict dynamic scapulothoracic lateral rotation accurately using skin-mounted scapula markers.