We studied the effects of medial pallidotomy in the first 20 consecutive patients with Parkinson's disease (PD) undergoing this MRI/electrophysiologically guided procedure at our institution. The mean age of patients was 65.5 years (median 66.5) and none suffered any serious complications. Pallidotomy significantly improved motor function in both "on" and "off" states as measured by Unified Parkinson's Disease Rating Scale (UPDRS) motor scores and timed tests (Purdue pegboard and counter tapping) in the arm contralateral to surgery 3 months postoperatively. Patients also improved in terms of activities of daily living, reflected by improved UPDRS activity of daily living and complications of therapy scoring and reduced levodopa-induced dyskinesias; six of 11 patients who could not walk in an "off" state prior to surgery could do so postoperatively. The total UPDRS score improved by 22% from preoperative values. The aforementioned improvements occurred similarly in patients greater than (n = 11) or less than 65 years (n = 9) at surgery. Neuropsychological measures indicated that although the majority of cognitive function remains unchanged in right-handed PD patients following dominant (left) hemisphere pallidotomy, mild specific declines in word generation are present. The findings of this study suggest that unilateral pallidotomy is safe and associated with improved motor functioning in elderly as well as younger PD patients experiencing significant disability despite optimal medical therapy.