The renaissance of functional neurosurgery in the treatment of Parkinson’s disease has sparked also the interest in other movement disorders which are refractory to medical treatment. Deep brain stimulation (DBS) has been used only since a few years in dystonia patients. This review summarizes the available data on pallidal and thalamic DBS for various dystonic syndromes. The major advantage of DBS as compared to radiofrequency lesioning is that it allows performing contemporaneous bilateral surgery with relatively low morbidity in these patients. The posteroventral lateral globus pallidus internus (GPi) has been the preferred target in most instances, thus far. While phasic dystonic movements may improve early after surgery, the response of tonic dystonic movements to chronic stimulation may be delayed. The most beneficial results have been achieved in patients with primary genetic generalized and segmental dystonia, myoclonic dystonia, and complex cervical dystonia. Outcome has been varied in patients with other dystonic disorders, in particular those with secondary dystonia. Most studies have reported on relatively short follow-up periods, on single cases, or were retrospective. Pallidal DBS has been shown to be effective in complex cervical dystonia yielding both symptomatic and functional benefit for up to 2.5 years of follow-up. Dramatic improvement has been obtained in children and in adults with DYT1 positive dystonia. Also, patients with non DYT1 genetic dystonia achieved sustained benefit for up to 2 years of follow-up. Preliminary experience indicates that choreoathetosis in patients with cerebral palsy responds less well to pallidal DBS, and that it may not be effective at all in some patients. In single instances unilateral pallidal DBS has been shown to yield valuable benefit in patients with hemidystonia. The experience with DBS for treatment of Meige syndrome and other focal dystonias has been explored only recently. There is much less experience with thalamic DBS for dystonia. Thalamic DBS has been shown to be effective in single cases with posttraumatic dystonia, postanoxic dystonia and paroxysmal nonkinesigenic dystonia. Future perspectives of DBS for treatment of dystonia include the development of new technology, the evaluation of the possible role of other targets, and carefully planned studies to further establish the role of surgery.