Background: Thyroidectomy as a first line treatment for Graves' disease is rarely utilized in the US. The purpose of this study was to analyze the safety and efficacy of thyroid surgery among patients with Graves' disease. Methods: Fifty-six patients with Graves' disease underwent thyroid surgery between May 1994 and May 2008 at a single academic institution. Preoperative, intraoperative, and postoperative variables were analyzed. Results: A total of 58 surgeries were performed: 55.1% (n = 32) total thyroidectomy, 41.3% (n = 24) subtotal/lobectomy, 3.4% (n = 2) completion thyroidectomy. The average gland weight was 47.3 ± 10.8 gm, with 70% weighing > 30 gm. Reasons for having thyroid surgery were persistent disease despite medical therapy (46.6%), patient preference (24.1%), multinodular goiter/cold nodules (20.3%), failed RAI (radioactive iodine) treatment (16%), and opthalmopathy (12.1%). Of those patients that failed prior RAI therapy, the only factor that was predictive of failure was disease severity, as demonstrated by a markedly elevated initial free-T4 value (11.8 ± 4.5 ng/dL, P = 0.04). Transient symptomatic hypocalcemia occurred in 10.7% (n = 6) of patients, and one patient (1.8%) had symptomatic hypocalcemia lasting > 6 mo. There were no permanent recurrent laryngeal nerve injuries. There was no difference in overall complication rates between patients based on surgical procedure (subtotal versus total thyroidectomy), preoperative RAI treatment, or gland size. Recurrences occurred in 6% of the subtotal thyroidectomy group and 0% of the total thyroidectomy group (P = 0.008). Conclusion: Thyroidectomy for patients with Graves' disease can be performed with very low complication rates and when a total thyroidectomy is performed, there is almost no risk of recurrence. © 2011 Elsevier Inc. All rights reserved.