For the integration with the environment, people grasp and manipulate objects with hands. The tasks are followed and controlled with eyes. The interaction between sensory information is a part of a complex control loop which must be determined. For the control loop, it is necessary to measure influences between the visual perception and the execution of the movements. In this case, a simple test of angular reproduction, as in gait analysis, is not enough because the 3D movements of the arm and the point of gaze shall be captured.A robot executes a defined trajectory and the subject must follow it with the finger. The movements of the upper extremity are captured with a motion analysis system and the angles of the wrist, elbow, shoulder and sternoclavicular joint are determined using a biomechanical model. At the same time, the subject uses an Eye-Tracker to capture the movements of the eyes during the execution of the robot's trajectory. The positions in relation to the robot are determined in space with the motion analysis system. Therefore, it is possible to define the point of gaze, the robot and the subject in a common coordinate system.The aim of this work is the development of a mathematical model based on eye-hand coordination described by a control loop whose inputs are the proprioception and the visual perception. The subject looks his finger and the movements of the robot and tries to follow the movements. The output of the control loop is the movement of the finger. The elements of the model shall be determined and therefore, the right choice of the variables and the movements of the robot are essential.With experiments of eye-hand coordination based on control loops, different simulations can be executed in the next steps. Therefore, dysfunctions related to the perception and proprioception can be investigated. The parameters of healthy subjects determine trajectories which are compared with trajectories of patients with sensory motor dysfunctions. The correct assumption of the control strategy used in the control loop shall allow the simulation of proprioception through the changes of the parameters and the inputs.For this reason, a pratical and flexible tool for the quantitative analysis in time and space between the free movements of the arm, head and eyes shall be available and the modelling of the control startegy based on natural eye-hand coordination must be inclusively the proprioceptive skills on the base of the whole system.