PURPOSE: Lymph node analysis is essential for staging gastric cancer. Intraoperative lymphatic mapping and sentinel lymphadenectomy have not yet been investigated for most gastrointestinal neoplasms. The purpose of this study is to evaluate the usefulness of ex vivo lymphatic mapping in patients with gastric cancer. METHODS: 42 patients with gastric cancer underwent ex vivo lymphatic mapping and sentinel lymph node (SN) biopsy after standard surgical resection from March 2002 to September 2002. Within 5 minutes of resection, stomach specimens were injected submucosally around the tumor with isosulfan blue dye. Blue lymphatic channels were identified and followed to the blue-stained SN (s) which were harvested. The specimen was fixed in formalin and subsequently analyzed in the usual fashion. RESULTS: At least one SN was identified in 39 patients (92.9%). The average number of SNs identified was 2.5 (range, 1~6), and the average number of nodes in each gastric cancer specimen was 23.4 (range, 13~55). 14 patients had lymph nodes containing metastatic disease. 9 patients had metastasis in both sentinel and nonsentinel node. In 5 patients the sentinel nodes was negative for disease, whereas the nonsentinel lymph nodes contained metastatic disease (false negative rate=35.7%). Of these 5 patient, one may have skip metastasis and four had metastasis on same rerional lymph node group. CONCLUSION: Ex vivo SN mapping of the stomach is technically feasible, but it is too early to provide a useful approach to evaluate lymph node metastasis.