During isokinetic shoulder elevation in the coronal and sagittal planes, the centre of rotation of the glenohumeral joint was displaced 8-cm vertically relative to the centre of rotation of the dynamometer's actuator arm. It was suggested that in order to account for this displacement, the recorded maximum moment values should be reduced by 12.5%. displacement of the glenohumeral centre of rotation (DGCR) during isokinetic shoulder elevation in the scapular plane and to conduct an error analysis of the dynamometer's moment recordings. The DGCR was assessed in twenty healthy volunteers, during isokinetic concentric elevation of the left shoulder, in the scapular plane, using a two-dimension motion analysis system. Measurements were performed at 30, 60 and 120°·s
$^{-1}$
, between 30--90° of shoulder elevation and were analysed every 12°. The glenohumeral centre of rotation was superomedially displaced relative to the centre of rotation of the actuator arm, throughout shoulder elevation (P < 0.001). The effect of isokinetic velocity on the DGCR was not significant. The DGCR resulted also in changes of humeral lever arm orientation and length relative to the pre-set orientation and length of lever arm. As a result of these changes, dynamometer's recordings underestimated corrected angle-based maximum moments by ≤ 6.2%. isokinetic elevation of the shoulder in the scapular plane was probably attributed to normal kinematics of shoulder elevation that requires synchronous function of both the scapulothoracic and glenohumeral joints. The cutaneous markers and the 2-D kinematic analysis system used for the assessment of a 3-D shoulder motion may effect the precision in detecting movements of the shoulder under the skin. The present findings are also limited to the position, stabilisation method and range of motion tested.
