This paper discusses the formation and control of optimal trajectories in human multijoint arm movements. The authors propose a mathematical model based on the minimum torque-change criterion, which suggests that the central nervous system (CNS) selects the optimal trajectory by minimizing the integral of the square of the rate of change of torque over the entire movement. This model is formulated by defining an objective function, which measures the performance of any possible movement. The objective function is defined as the integral of the square of the rate of change of torque over the entire movement. The model is then used to evaluate human hand trajectories under various behavioral situations, supporting the idea that the human hand trajectory is planned and controlled according to the minimum torque-change criterion. The paper also presents a computational model of voluntary movement, which accounts for Marr's first level of understanding complex information processing systems. The model proposes that three computational problems are solved at different levels in the CNS: (1) determination of a desired trajectory, (2) transformation of visual coordinates of the desired trajectory to body coordinates, and (3) generation of motor commands to realize the desired trajectory. The model suggests that there are other information processing steps that realize the desired trajectory, which are not shown in the step-by-step information processing hypothesis. The paper discusses the kinematic and dynamic aspects of multijoint arm movements, noting that they are more complex than single-joint movements due to the presence of interactional forces. The authors also discuss the strategy used by the CNS to determine a desired trajectory and the coordinates frame in which the trajectory is planned. The results of experiments suggest that the desired trajectory is first planned at the task-oriented (visual) coordinates, and that the hand kinematics are used to plan arm movements rather than joint rotations. The paper concludes that the minimum torque-change criterion is a key factor in the formation and control of optimal trajectories in human multijoint arm movements.This paper discusses the formation and control of optimal trajectories in human multijoint arm movements. The authors propose a mathematical model based on the minimum torque-change criterion, which suggests that the central nervous system (CNS) selects the optimal trajectory by minimizing the integral of the square of the rate of change of torque over the entire movement. This model is formulated by defining an objective function, which measures the performance of any possible movement. The objective function is defined as the integral of the square of the rate of change of torque over the entire movement. The model is then used to evaluate human hand trajectories under various behavioral situations, supporting the idea that the human hand trajectory is planned and controlled according to the minimum torque-change criterion. The paper also presents a computational model of voluntary movement, which accounts for Marr's first level of understanding complex information processing systems. The model proposes that three computational problems are solved at different levels in the CNS: (1) determination of a desired trajectory, (2) transformation of visual coordinates of the desired trajectory to body coordinates, and (3) generation of motor commands to realize the desired trajectory. The model suggests that there are other information processing steps that realize the desired trajectory, which are not shown in the step-by-step information processing hypothesis. The paper discusses the kinematic and dynamic aspects of multijoint arm movements, noting that they are more complex than single-joint movements due to the presence of interactional forces. The authors also discuss the strategy used by the CNS to determine a desired trajectory and the coordinates frame in which the trajectory is planned. The results of experiments suggest that the desired trajectory is first planned at the task-oriented (visual) coordinates, and that the hand kinematics are used to plan arm movements rather than joint rotations. The paper concludes that the minimum torque-change criterion is a key factor in the formation and control of optimal trajectories in human multijoint arm movements.