Since the 1960s, deep brain stimulation and spinal cord stimulation at low frequency (30 Hz) have been used to treat intractable pain of various origins. For this purpose, specific hardware have been designed, including deep brain electrodes, extensions, and implantable programmable generators (IPGs). In the meantime, movement disorders, and particularly parkinsonian and essential tremors, were treated by electrolytic or mechanic lesions in various targets of the basal ganglia, particularly in the thalamus and in the internal pallidum. The advent in the 1960s of levodopa, as well as the side effects and complications of ablative surgery (e.g., thalamotomy and pallidotomy), has sent functional neurosurgery of movement disorders to oblivion. In 1987, the serendipitous discovery of the effect of high-frequency stimulation (HFS), mimicking lesions, allowed the revival of the surgery of movement disorders by stimulation of the thalamus, which treated tremors with limited morbidity, and adaptable and reversible results. The stability along time of these effects allowed extending it to new targets suggested by basic research in monkeys. The HFS of the subthalamic nucleus (STN) has profoundly challenged the practice of functional surgery as the effect on the triad of dopaminergic symptoms was very significant, allowing to decrease the drug dosage and therefore a decrease of their complications, the levodopa-induced dyskinesias. In the meantime, based on the results of previous basic research in various fields, HFS has been progressively extended to potentially treat epilepsy and, more recently, psychiatric disorders, such as obsessive-compulsive disorders, Gilles de la Tourette tics, and severe depression. Similarly, suggested by the observation of changes in PET scan, applications have been extended to cluster headaches by stimulation of the posterior hypothalamus and even more recently, to obesity and drug addiction. In the field of movement disorders, it has become clear that STN stimulation is not efficient on the nondopaminergic symptoms such as freezing of gait. Based on experimental data obtained in MPTP-treated parkinsonian monkeys, the pedunculopontine nucleus has been used as a new target, and as suggested by the animal research results, its use indeed improves walking and stability when stimulation is performed at low frequency (25 Hz). The concept of simultaneous stimulation of multiple targets eventually at low or high frequency, and that of several electrodes in one target, is being accepted to increase the efficiency. This leads to and is being facilitated by the development of new hardware (multiple-channel IPGs, specific electrodes, rechargeable batteries). Still additional efforts are needed at the level of the stimulation paradigm and in the waveform. The recent development of nanotechnologies allows the design of totally new systems expanding the field of deep brain stimulation. These new techniques will make it possible to not only inhibit or excite deep brain structures to alleviate abnormal symptoms but also open the field for the use of recording cortical activities to drive neuroprostheses through brain-computer interfaces. The new field of compensation of deficits will then become part of the field of functional neurosurgery.