AIM: To analyze the diagnostic utility of a small-caliber endoscope(SC-E) and clinicopathological features of false-negative gastric cancers(FN-GCs). METHODS: A total of 21638 esophagogastroduodenoscopy(EGD) gastric cancer(GC) screening examinations were analyzed. Secondary endoscopic examinations(n = 3352) were excluded because most secondary examinations tended to be included in the conventional endoscopy(C-E) group. Detection rates of GCs and FN-GCs were compared between SC-E and C-E groups. FN-GC was defined as GC performed with EGD within the past 3 years without GC detection. Macroscopic types, histopathological characteristics and locations of FN-GCs were compared with firstly foundgastric cancers(FF-GCs) in detail. RESULTS: SC-E cases(n = 6657) and C-E cases(n = 11644), a total of 18301 cases, were analyzed. GCs were detected in 16(0.24%) SC-E cases and 40 C-E(0.34%) cases(P = 0.23) and there were 4 FN-GCs(0.06%) in SC-E and 13(0.11%) in C-E(P = 0.27), with no significant difference. FN-GCs/GCs ratio between SC-E and C-E groups was not significantly different(P = 0.75). The comparison of endoscopic macroscopic types of FN-GCs tended to be a less advanced type(P = 0.02). Histopathologically, 70.6% of FN-GCs were differentiated and 29.4% undifferentiated type. On the other hand, 43.0% of FF-GCs were differentiated and 53.8% undifferentiated type, so FN-GCs tended to be more differentiated type(P = 0.048). CONCLUSION: The diagnostic utility of SC-E for the detection of GCs and FN-GCs was not inferior to that of C-E. Careful observation for superficially depressed type lesions in the upper lesser curvature region is needed to decrease FN-GCs. AIM: To analyze the diagnostic utility of a small-caliber endoscope (SC-E) and clinicopathological features of false-negative gastric cancers (FN-GCs). METHODS: A total of 21638 esophagogastroduodenoscopy (EGD) Secondary endoscopic examinations (n ​​= 3352) were excluded because most examinations tended to be included in the conventional endoscopy (CE) group. Detection rates of GCs and FN-GCs were compared between SC-E and CE groups. FN- GC was defined as GC performed with EGD within the past 3 years without GC detection. Macroscopic types, histopathological characteristics and locations of FN-GCs were compared to first found gastric cancers (FF-GCs) in detail. RESULTS: SC-E cases (n = 6657) and CE cases (n = 11644), a total of 18301 cases, were analyzed. GCs were detected in 16 (0.24%) SC-E cases and 40 CE 4 FN-GCs (0.06%) in SC-E and 13 (0.11%) in CE (P = 0.27) with no significant diffe The comparison of the endoscopic macroscopic types of FN-GCs tended to be less advanced type (P = 0.02). Histopathologically, The percentages of FN-GCs were differentiated and 29.4% undifferentiated type. On the other hand, 43.0% of FF-GCs were differentiated and 53.8% undifferentiated type, so FN-GCs tended to be more differentiated type (P = 0.048). CONCLUSION : The diagnostic utility of SC-E for the detection of GCs and FN-GCs was not inferior to that of CE. Careful observation for superficially depressed type lesions in the upper lesser curvature region is needed to decrease FN-GCs.