Many studies have focused on the analysis of risk factors for lymph node metastasis in early gastric cancer, but few studies analyze the drainage patterns of metastatic lymph nodes.Data was retrospectively analyzed from a database of gastric cancer resections from 2014-2018. The cohort included 786 pT1 patients with complete data. Outcomes evaluated were lymph node metastasis frequencies, survival analyses, and risk factors impacting prognosis. Lymph node mapping was conducted to explore the distribution pattern of metastatic nodes based on tumor location.The overall lymph node metastasis rate was 23.7%. Lymph node mapping showed that No. 7 (67.7%) and No. 3 (45.7%) group had the highest metastasis rates and regardless of tumor location, metastases were concentrated in perigastric lymph node groups (No. 3, 4, 5, 6, and 7), while involvement of second-tier station nodes was rare. The 5-year overall survival rate (54.8% vs 95.7%; P < 0.001) and disease-free survival rate (48.4% vs 95.7%; P < 0.001) of patients with node-positive disease were significantly worse than those of patients with node-negative disease. Multivariable Cox regression identified tumor size > 2cm (P = 0.007, < 0.001), poor differentiation (P = 0.007, < 0.001), T1b stage (all P < 0.001), lymph node metastasis (all P < 0.001), and vascular invasion (all P = 0.002, 0.016) as independent negative prognostic factors affecting 5-year OS and DFS in patients with early gastric cancer. Postoperative chemotherapy (P < 0.001, 0.019）was an independent positive prognostic factor.This retrospective study demonstrates that lymph node metastasis in early gastric cancer is widely and disorderly not depending on the location. Lymph node metastasis mainly occurs in the perigastric station, supporting a specific lymph node dissection strategy may be sufficient for some early gastric cancer. Meanwhile its prognosis is closely related to lymph node metastasis.The participant data with identifiers used to support the findings of this study were supplied by Qun Zhao under license and so cannot be made freely available. Requests for access to these data should be made to Qun Zhao, zhaoqun@hebmu.edu.cn.Copyright © 2025. Published by Elsevier Inc.