Background and Aims: To confirm the utility of a newly- devised endoscopic ultrasonography (EUS) echogenicity classification system, which is based on the normal five- layer structure of the gastrointestinal wall, for the differential diagnosis of gastrointestinal neoplasms. Methods: A total of 275 patients with gastrointestinal disease detected by endoscopy, 53 with submucosal tumor (22 esophageal, 27 gastric, and 4 duodenal), 208 with cancer (58 esopha- geal, 150 gastric), and 14 with gastric malignant lymphoma, were examined. In each case, we performed standard EUS with 7.5, 12, or 20 MHz to evaluate the echogenicity of the lesion. Echogenicity was classified into seven levels that is based on the echogenicity of the normal five- layer structure of the gastrointestinal wall. The levels are stratified from anechoic (level 0) to hyperechoic (level 6), muscularis propria and deep mucosa being referred to as level 2 and level 4, respectively. Level 6 was equivalent to or higher than the echogenicity of submucosa. Results: EUS findings revealed several distinct echogenicities among the diseases. In submucosal tumor, levels 0, 2, and 6 were observed only in patients with cyst, myogenic tumor, and lipoma, respectively. In malignancies, levels 0, 5, and 6 were not observed in this study. The echo level of malignant lymphoma was markedly lower than that of epithelial cancer. Conclusion: Our classification system of echogenicity can be a useful method in the differential diagnosis of gastrointestinal neoplasms. Background and Aims: To confirm the utility of a newly-devised endoscopic ultrasonography (EUS) echogenicity classification system, which is based on the normal five-layer structure of the gastrointestinal wall, for the differential diagnosis of gastrointestinal neoplasms. Methods: A total of 275 with gastrointestinal disease detected by endoscopy, 53 with submucosal tumor (22 esophageal, 27 gastric, and 4 duodenal), 208 with cancer (58 esopha- geal, 150 gastric), and 14 with gastric malignant lymphoma, were examined. In each case, we performed standard EUS with 7.5, 12, or 20 MHz to evaluate the echogenicity of the lesion. Echogenicity was classified into seven levels that is based on the echogenicity of the normal five- layer structure of the gastrointestinal wall. The levels are stratified from anechoic (level 0) to hyperechoic (level 6), muscularis propria and deep mucosa being referred to as level 2 and level 4, respectively. Level 6 was equivalent to or higher than the In submucosal tumor, levels 0, 2, and 6 were observed only in patients with cyst, myogenic tumor, and lipoma, respectively. In malignancies, levels 0, 5 , and 6 were not observed in this study. The echo level of malignant lymphoma was markedly lower than that of epithelial cancer. Conclusion: Our classification system of echogenicity can be a useful method in the differential diagnosis of gastrointestinal neoplasms.