The authors report the results of 782 microvascular decompression procedures for hemifacial spasm in 703 patients (705 sides), with follow-up study from 1 to 20 years (mean 8 years). Of 648 patients who had not undergone prior intracranial procedures for hemifacial spasm, 65% were women; their mean age was 52 years, and the mean preoperative duration of symptoms was 7 years. The onset of symptoms was typical in 92% and atypical in 8%. An additional 57 patients who had undergone prior microvascular decompression elsewhere were analyzed as a separate group. Patients were followed prospectively with annual questionnaires. Kaplan-Meier methods showed that among patients without prior microvascular decompression elsewhere, 84% had excellent results and 7% had partial success 10 years postoperatively. Subgroup analyses (Cox proportional hazards model) showed that men had better results than women, and patients with typical onset of symptoms had better results than those with atypical onset. Nearly all failures occurred within 24 months of operation; 9% of patients underwent reoperation for recurrent symptoms. Second microvascular decompression procedures were less successful, whether the first procedure was performed at Presbyterian-University Hospital or elsewhere, unless the procedure was performed within 30 days after the first microvascular decompression. Patient age, side and preoperative duration of symptoms, history of Bell's palsy, preoperative presence of facial weakness or synkinesis, and implant material used had no influence on postoperative results. Complications after the first microvascular decompression for hemifacial spasm included ipsilateral deaf ear in 2.6% and ipsilateral permanent, severe facial weakness in 0.9% of patients. Complications were more frequent in reoperated patients. In all, one operative death (0.1%) and two brainstem infarctions (0.3%) occurred. Microvascular decompression is a safe and definitive treatment for hemifacial spasm with proven long-term efficacy.