Background: Adenoma with high-grade dysplasia (HGD) is defined as carcinoma that has not breached the muscularis mucosa. Adenomas with HGD are considered cured with polypectomy. However, endoscopic polypectomy can be technically challenging given larger sizes and greater risk of inadequate resection and thus recurrence. The alternative is surgical resection. The purpose of this study is to compare the recurrence rate of HGD or colorectal carcinoma (CRC) and complications in patients who were managed with endoscopic polypectomy versus those who were treated surgically. Methods: Patients at the Michael E. DeBakey VAMC undergoing colonoscopy from 2005-2010 with adenomas with HGD were identified and categorized by management approach (endoscopy vs surgery). Patients who underwent only endoscopic polypectomywere compared against those who underwent surgical resection based on patient demographics, polyp characteristics, and two major endpoints: recurrence of HGD/CRC and complication rates. Results: One hundred ninety four patients with HGD were identified at index colonoscopy. Of those, 161 underwent only endoscopic polypectomy and 33 underwent surgical resection. There were no significant differences in baseline patient characteristics (age, comorbidity, race, BMI, number of polyps, proportion FOBT/ FIT positive). However, polyps in the surgery group tended to be larger (mean of 2.0 vs 3.6 cm, p<0.0001) and to have piecemeal resection at index colonoscopy (26.7% vs 76.5%, p<0.0001). Among 161 endoscopic patients, over a median follow up of 2.0 years, 7.4% developed an adenoma with HGD or CRC (5.6% and 1.9%, respectively) with a 0.6 % complication rate (bleeding). All had documented visual clearance of adenoma at prior colonoscopy. Thirteen of 161 (8.0%) patients died during follow up, of which one (0.6%) was attributed to CRC. Among 33 surgical patients, one (3.0%) developed an adenoma with HGD and none developed interval CRC over a median follow up of 2.4 years. However, the surgical group had a complication rate of 15.2% (wound infection, renal failure, bowel obstruction and one post-operative death from sepsis). Eight of the 33 (32%) patients died during follow up, none of which was attributed to CRC. The complication rate for endoscopic polypectomy was significantly lower than that of surgical resection (0.6% vs 15.2%, p= 0.0004), while the recurrence rate was not different (7.4% vs 3.0%, p=0.7). Conclusion: Endoscopic resection of polyps with HGD seems to be as effective as surgery, but with lower complication rates. Balancing risks and benefits between surgical and endoscopic therapy is difficult, but attempts at repeat endoscopy for complete resection of HGD should be attempted before surgical resection, especially in an elderly population with significant comorbidities.