Microscopic examination of the proximal and distal resection margins is part of the routine pathologic evaluation of colorectal surgical specimens removed for adenocarcinoma. Anastomotic donuts are frequently received and microscopically examined. We examined 594 specimens received over a period of 10 years and found only 3 cases of definitive direct involvement of a longitudinal margin by carcinoma. All 3 cases also showed tumor at the margin grossly. One case of margin involvement by adenocarcinoma was found in which the tumor was grossly 1.7 cm away; however, this finding was likely a tumor deposit, as the patient had diffuse metastatic disease. All 242 anastomotic donuts examined were free of carcinoma. Our study suggests that the proximal and distal margins of colorectal cancer specimens need not be examined microscopically in order to accurately assess margin status in cases where the tumor is at least 2 cm away from the margin of resection. Also, in cases in which anastomotic donuts are included with the case, these should be considered the true margins of resection and may be microscopically examined in place of the bowel specimen margins when margin examination is needed. Anastomotic donuts need not be examined if the tumor is more than 2 cm away from the margin. An exception to this rule would be cases of rectal adenocarcinoma where neoadjuvant therapy is given prior to surgery. In these cases, mucosal evidence of malignancy may be absent and microscopic examination of the margins is the only way to assure complete excision.