Background:Hand synergies have been extensively studied over the last few decades. Objectives of such researchare numerous. In neuroscience, the aim is to improve the understanding of motor control and its ability to reduce thecontrol dimensionality. In applied research fields like robotics the aim is to build biomimetic hand structures, or inprosthetics to design more performant underactuated replacement hands. Nevertheless, most of the synergyschemes identified to this day have been obtained from grasping experiments performed with one single (generallydominant) hand to objects placed in a given position and orientation in space. Aiming at identifying more genericsynergies, we conducted similar experiments on postural synergy identification during bimanual manipulation ofvarious objects in order to avoid the factors due to the extrinsic spatial position of the objects.Methods:Ten healthy naive subjects were asked to perform a selected “grasp-give-receive” task with both handsusing 9 objects. Subjects were wearing Cyberglove©on both hands, allowing a measurement of the joint posture (15degrees of freedom) of each hand. Postural synergies were then evaluated through Principal Component Analysis(PCA). Matches between the identified Principal Components and the human hand joints were analyzed thanks to thecorrelation matrix. Finally, statistical analysis was performed on the data in order to evaluate the effect of some specificvariables on the hand synergies: object shape, hand side (i.e., laterality) and role (giving or receiving hand).Results:Results on PCs are consistent with previous literature showing that a few principal components might besufficient to describe a large variety of different grasps. Nevertheless some simple and strong correlations betweenPCs and clearly identified sets of hand joints were obtained in this study. In addition, these groupings of DoFcorresponds to well-defined anatomo-functional finger joints according to muscle groups. Moreover, despite ourprotocol encouraging symmetric grasping, some right-left side differences were observed.Conclusion:The set of identified synergies presented here should be more representative of hand synergies ingeneral since they are based on both hands motion. Preliminary results, that should be deepened, also highlight theinfluence of hand dominance and side. Thanks to their strong correlation with anatomo-functional joints, thesesynergies could therefore be used to design underactuated robotics hands.