INTRODUCTION:The phenotype of hereditary nonpolyposis colorectal cancer includes an 80 percent lifetime risk of colorectal cancer, a predominance of lesions proximal to the splenic flexure, and a high incidence of synchronous and metachronous neoplasia. Although prophylactic colectomy is rarely advised for patients with a hereditary nonpolyposis colorectal cancer genotype and a normal colon, the presence of advanced neoplasia in the context of a qualifying family history or a hereditary nonpolyposis colorectal cancer genotype has led to such recommendations. We performed this study to document the patterns of colorectal surgery performed for cancer‐bearing patients who are part of an Amsterdam criteria‐positive family and to compare rates of metachronous cancers that follow index total or segmental colectomy.METHODS:Family trees fulfilling the classic Amsterdam criteria for hereditary nonpolyposis colorectal cancer were identified, and all patients for whom surgical and pathology records were available were included in the study. Type of surgery and the outcome of subsequent follow‐up were abstracted. Patients were divided into those treated at the Cleveland Clinic and those treated elsewhere.RESULTS:There were 39 families with 93 affected patients. These patients had 127 colorectal cancers, 76 (60 percent) of which were right sided (proximal to the splenic flexure). Median age at diagnosis of the index cancer was 47 (range, 26‐81) years. Sixteen patients (17 percent) had metachronous cancers and multiple surgeries, whereas four (4 percent) had synchronous cancers. Median follow‐up for patients who underwent surgery at the Cleveland Clinic was 13 (range, 4‐49) years, whereas that for those who underwent surgery elsewhere was 14 (range, 1‐42) years. Sixteen (48 percent) of the 33 patients who underwent surgery at the Cleveland Clinic had a total colectomyvs.7 (12 percent) of the 60 who had surgery elsewhere (Fisher's exact test,P< 0.001). Only one patient who had surgery at the Cleveland Clinic had a second operation for a metachronous cancer (1/17 patients having a segmental resection). Fifteen patients who underwent surgery elsewhere needed a second resection for metachronous cancer (15/53 patients having a segmental resection; Fisher's exact test,P= 0.094).CONCLUSION:We conclude that there is high risk of metachronous colorectal cancer if an index cancer in a hereditary nonpolyposis colorectal cancer patient (defined according to Amsterdam criteria) is treated by partial colectomy. However, this risk can be lowered by performing a total colectomy at the time of index surgery, or possibly by effective postoperative surveillance.