Background: Endoscopic treatments, such as endoscopic submucosal dissection (ESD) and laparoscopic gastrectomy, are increasingly used to treat a subset of patients with early gastric cancer (EGC). To achieve successful outcomes, it is very important to accurately determine the lateral extent of the tumor. Therefore, we investigated the diagnostic performance of chromoendoscopy using indigo carmine dye added to acetic acid (AI chromoendoscopy) in delineating differentiated or undifferentiated adenocarcinomas in patients with EGC. Methods: We prospectively included 151 lesions of 141 patients that had an endoscopic diagnosis of EGC. All the lesions were examined by conventional endoscopy and AI chromoendoscopy before ESD or laparoscopic gastrectomy. The border clarification between the lesion and the normal mucosa was classified as distinct or indistinct before and after AI chromoendoscopy. Results: The borders of the lesions were distinct in 66.9% (101/151) with conventional endoscopy and in 84.1% (127/151) with AI chromoendoscopy (P < 0.001). Compared with conventional endoscopy, AI chromoendoscopy clarified the border in a significantly higher percentage of differentiated adenocarcinomas (74/108 [68.5%] vs 97/108 [89.8%], respectively, P < 0.001). However, the border clarification rate for undifferentiated adenocarcinomas did not differ between conventional endoscopy and AI chromoendoscopy (27/43 [62.8%] vs 30/43 [70.0%], respectively, P = 0.494). Conclusions: AI chromoendoscopy is useful in determining the lateral extent of EGCs. However, its usefulness is reduced in undifferentiated adenocarcinomas.