The role of surgical treatment for dystonia is evolving. Past reports of pallidotomies and thalamotomies have been mixed, with remarkable success and outright failures being reported in the same paper, ofttimes for apparently similar types of patients. Reports in the literature do little to clarify why some patients with dystonia improve and others do not. In general, it would appear that thalamotomy or pallidotomy is best performed for focal, segmental, or hemidystonic symptoms involving appendicular musculature and that bilateral procedures give the best results for generalized or axial dystonia and, depending on the musculature involved, for patients with spasmodic torticollis. Nevertheless, bilateral thalamotomy is associated with an increased risk of side effects, and bilateral pallidotomy, although used effectively without complications, has not been explored in a large enough number of patients to adequately assess the relative incidence of side effects. Although a variety of variables have been considered predictive of outcome, including age, etiology, somatotopic distribution, and associated symptomatology, there has been little advance in the understanding of the importance of these factors over the last decades. A number of factors complicate this research, including the variety of target sites chosen for lesioning, the lack of systematic and standardized methods for patient assessment, the heterogeneous patient population, and the lack of verification of lesion location. More recently, however, armed with an improved understanding of the pathophysiology underlying the development of hypokinetic and hyperkinetic movement disorders and with improved methods of target localization, contemporary reports of thalamotomy and pallidotomy for dystonia are more promising. In addition, the development of DBS offers an alternative to ablation in the surgical treatment of dystonia; although only a few cases of DBS for dystonia have been reported, results are promising. There remain, however, many pertinent questions concerning the role of surgical therapy for dystonia. What is the relative effect of thalamotomy versus pallidotomy? What is the optimal lesion location and size, and will this vary depending on etiology, age, or distribution of the dystonia? What are the predictors of clinical outcome? Answers to these questions are critical to the future development and improvement of surgical therapies for dystonia, and require well thought out, carefully designed clinical studies.