Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)Purpose: This study tested the hypothesis that muscle and interaction torques can be altered independently in order to improve in specific kinematics performance observed following practice. We also tested the hypothesis that a simple set of rules of EMG-control and kinetic-control models could explain the EMG and kinetic changes due to practice of movements with reversal. Scope: Kinematics of the upper arm with reversal, performed over three distances, was reconstructed using motion analysis. The muscle and interaction torques were calculated using inverse-dynamics. EMG activities of the major arm muscles were also recorded. The results demonstrate that improved performance is facilitated by an increase in muscle torque (and therefore acceleration) at the proximal joint (shoulder) and by an increase in the interaction torque at the distal joint (elbow). No changes were observed in the amount of muscle activity underlying these kinetic modifications, except for a decrease in the shoulder antagonist latency. Conclusion: The results confirm Bernstein's idea that the central nervous system takes advantage of the passive-interactive properties of the moving system. Also the modulation of the EMG patterns should be explained taking in account the reactive forces and the dual functions (maintenance of posture and generation of movement) of the muscles. (C) 2007 Elsevier Ltd. All rights reserved.185771779Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)FAPESP [00/10527-6]