The focus of this review is the role of extended resection in the initial treatment of primary colorectal carcinoma. About 10% of patients with newly diagnosed colorectal cancer will have locally advanced disease without evident distant or discontiguous intraabdominal metastases. En bloc resection of such tumors, including attached tissues or organs, provides a 5-year survival rate of about 40%, if the microscopic margins are tumor free. As many as 60% of these large tumors are node negative; in this circumstance the 5-year survival rate approaches 70%. These results are achievable when there is a meticulous preoperative and intraoperative search for metastases, a wide anatomic resection, including en bloc lymphadenectomy, is performed, and tumor manipulation is minimized. Blunt separation of structures adherent to the primary tumor should be avoided because adhesions will be neoplastic in about 50% of cases, and cancer recurrence is virtually certain when tumor is transected. The mortality from multivisceral resection, including total pelvic exenteration, should be 10% or less. We emphasize the importance of including these patients in prospective trials to define their optimal adjuvant therapy. There is a disturbing recurring theme in published series, failure to extend the scope of resection in potentially curable patients. The management of these locally advanced lesions typically receives but cursory notice in otherwise highly detailed reviews or textbook chapters. In the present era of emerging multimodality treatment for colorectal cancer, the adequacy of the one most important treatment component--surgical resection--is seldom emphasized. Furthermore, our perusal of the recent literature disclosed no diminution in the incidence of inadequate resection, suggesting that this subject requires more emphasis in postgraduate surgical education.