Fast aiming movements were measured in a choice reaction paradigm in a healthy control group and in Parkinsonian patients. The patients were tested without (‘off ’) and with 3,4-dihydroxyphenylalanine (‘on’) (L-dopa) medication. The movement trajectories were used to estimate the parameters of a dynamic linear model. The model is based on the functional structure of the basal ganglia-thalamocortical circuit with direct and indirect pathways linking the putamen to the basal ganglia output nuclei (Albin et al. 1989). The output of the circuit is connected to a model for the motor neuron-musculo-skeletal system. The gain k _d for the direct pathway and the gain k _i for the indirect pathway were estimated. They were found to be significantly decreased for Parkinsonian patients in ‘off ’ compared with the control group. L-dopa therapy in Parkinsonian patients increased the gains of the direct and the indirect pathway almost to normal values which implies that the long-term dopamine level in the striatum was excitatory for the direct and for the indirect pathway. This result is restricted to movements of correct size. For movements of diminished size, which are typical for Parkinsonian patients, the model predicts that the dopamine level in the striatum is excitatory for the direct pathway but inhibitory for the indirect pathway. The simulated values for neuronal activities are in agreement with expected values according to the experimental data. The proposed model of the ‘motor’ basal ganglia thalamocortical circuit implies that information about biomechanical properties of the musculo-skeletal system is stored in the ‘motor’ basal ganglia-thalamocortical circuit, and that the basal ganglia are involved in computation of the desired movement amplitude.