Objective: To distinguish the endoscopic and clinical features of ampullary polyps,to investigate the endoscopic cancer risk factors of ampullary polyps based on the compared differences of benign lesions and adenocarcinoma, and to assess the accuracy of forceps biopsy.Methods: Authors retrospectively analyzed the data extracted from patients treated with endoscopic papillectomy (EP) from January 2009 to May 2019 in the Department of Gastroenterology, Peking University Third Hospital. Endoscopic pictures and pathology reports were reevaluated and analyzed. Differences between benign and cancer groups were conducted.Results: In the study,42 cases were involved, 35 to 83 years old, containing 83.3% older than 50 years old patients. The histological types were as follows, 2 for inflammatory polyps (4.8%), 1 for neuroendocrine tumor (2.4%), 1 for hyperplastic polyp (2.4%), 5 for grade Ⅰ adenoma (11.9%), 10 for grade Ⅱ adenoma (23.8%), 4 for grade Ⅲ adenoma (9.5%) and 19 for adenocarcinoma (45.2%),and 90.5% were adenoma or adenocarcinoma. The average age of benign group (inflammatory polyps and adenomas) was (56.7±9.2), which was significantly younger than that of adenocarcinoma group [(66.0±9.8), P=0.004]. Tumor diameter in adenocarcinoma group[(2.3±0.8) cm] was significantly larger than that in benign group[(1.6±0.6) cm, P=0.002]. Benign lesions only showed Yamada type Ⅰ(57.1%)and type Ⅱ(42.9%). The percentage of Yamada type Ⅰ (36.8%)and type Ⅱ(31.6%) in adenocarcinoma group was lower than that in benign group. Moreover, Yamada type Ⅲ (31.5%) was only found in the adenocarcinoma group. Significant differences were observed between the two groups in Yamada types (P=0.046). Most of the benign lesions had clear boundary(18/21,85.7%). The percentage of clear boundary in adenocarcinoma group (2/19,10.5%) was significantly lower than that in the benign group (P<0.001). No significant differences were investigated in color (P=0.353) and surface (P=0.324) between benign and adenocarcinoma lesions. Pooling age, lesion diameter, Yamada type and clear boundary into Logistic regression analysis, only age (OR=1.186, 95%CI 1.025-1.373, P=0.022) and clear boundary (OR=66.218, 95%CI 3.421-1 281.840, P=0.006) were the independent cancer risk factors. Only 2 (10.5%) in the 19 cancer patients had positive biopsy results before EP. As compared with post-EP, 55.3% (21/38) biopsies were under-estimated, including 17 (17/19, 89.5%) adenocarcinomas and 4 (4/10, 40%) grade Ⅱ adenomas.Conclusion: adenoma and adenocarcinoma were the major histological type of ampullary po-lyps. Age and unclear boundary were the independent risk factors of ampullary adenocarcinoma. Forceps biopsy was not enough for ampullary polyp differentiation.