There is little debate that deep brain stimulation (DBS) has been an effective tool in the treatment of Parkinson's disease as well as other movement disorders. There remains however, considerable debate concerning the mechanism(s) underlying its beneficial effect. The comparable effect of stimulation to ablation in the thalamus on tremor, and in the subthalamic nucleus (STN) and internal segment of the globus pallidus (GPi) on the motor signs associated with PD, have led many investigators to conclude that DBS acts to suppress neuronal activity, decreasing output from the stimulated site. There are, however, data that do not support this argument. Microdialysis studies in GPi showed increased levels of glutamate during STN stimulation, suggesting activation of glutamatergic output from the STN to the GPi. Studies in parkinsonian primates have demonstrated increased mean discharge rates of neurons in GPi during chronic stimulation in STN, and GPi stimulation in humans has been associated with a suppression of neuronal activity in the thalamus. Contrary to what one would expect if stimulation inhibits output from the stimulated structure, stimulation in GPe has been demonstrated to improve bradykinesia. Although arguments for increased output from the stimulated structure seem to conflict with the hypothesis that stimulation acts to inhibit neuronal activity, it is possible to explain these observations through a common mechanism, e.g. activation of fiber pathways. Based on this mechanism, the effect of stimulation on cellular activity in the stimulated site would be increased or decreased dependent on the neurotransmitter of the afferent fibers projecting to that site. However, in addition to activation of afferent fibers, projection axons from neurons in the stimulated structure, also readily excitable by electrical stimulation, would also be tonically activated and discharge independently of the soma, thereby increasing output from the structure during extracellular stimulation. Thus, although high frequency stimulation may inhibit neurons via activation of inhibitory afferents, the output from that structure may be increased as the result of activation of axonal elements leaving the target structure. This hypothesis would explain the present experimental results, is consistent with excitability profiles of neuronal elements based on their biophysical properties, and fits with more recent models emphasizing the role of altered patterns of neuronal activity in the development of hypokinetic and hyperkinetic movement disorders.