INTRODUCTION: In laparoscopic gastrectomy, a method to locate the margin of an early-stage cancerous lesion that is invisible from the serosal surface and impalpable during laparoscopic procedures is needed to determine an appropriate transection line. We conducted a prospective study to develop a new marking method using preoperative submucosal injection of indocyanine green (ICG). METHODS: Patients undergoing laparoscopic gastrectomy for T1 gastric cancer were recruited. The first 11 patients comprised the learning set and the subsequent 18 patients the validation set. ICG was endoscopically injected in the submucosal layer of the stomach approximately 1 cm away from the tumor edge 1 or 3 days before surgery. The diameters of the visualized ICG were compared with those of a conventional marking method using India ink in 10 historical controls. RESULTS: In the learning set, the optimal amount of ICG was determined to be 0.1 mL at a concentration of 0.5 mg/mL. In the validation set, the same procedure was repeated. No technical problems or adverse reactions related to ICG injection were observed. In all cases, ICG was successfully detected, and negative surgical margins were pathologically confirmed. The mean long diameter of the visualized ICG fluorescence measured at the mucosal surface of the stomach was significantly smaller in the current study than in the historical controls in whom India ink was used (21 vs 52 mm, P < 0.0001). CONCLUSIONS: The preoperative submucosal ICG marking was safely performed and successfully detected without excessive blurring during laparoscopic gastrectomy.