Object <p>The aim of this study was to elucidate the disparities in long-term survival outcomes between endoscopic therapy and gastrectomy in T1b gastric cancer (GC) patients</p> Methods <p>Data pertaining to T1b GC patients from 2010-2020 were extracted from the Surveillance, Epidemiology, and End Results (SEER) database. Additionally, clinical characteristics and survival metrics for pertinent patients were collated from the Chinese multicenter database. Overall survival (OS) and cancer-specific survival (CSS) were evaluated using various methodologies [unadjusted, stable inverse probability-weighted treatment (IPTW), and propensity score matching (PSM)] among the endoscopy, gastrectomy, and chemoradiotherapy cohorts. Variable selection was conducted through the Least Absolute Shrinkage and Selection Operator (Lasso) regression, culminating in the development and validation of a prognostic model across distinct cohorts</p> Results <p>In all, 2020 GC patients were identified through the SEER database. Multiple comparative frameworks were employed to scrutinize the differences in outcomes among gastrectomy, endoscopy, and chemoradiotherapy modalities. The analyses revealed no statistically significant disparity in long-term survival between gastrectomy and endoscopy while demonstrating that the efficacy of chemoradiotherapy was markedly inferior to that of endoscopy. Further validation was achieved by analyzing data from 116 patients from the Chinese multicenter database, corroborating the aforementioned findings. Subsequently, Lasso regression facilitated the variable selection for the prognostic model, which ultimately included race, gender, age, histologic grade, tumor dimensions, and treatment strategy. The model demonstrated satisfactory predictive accuracy for OS among T1b GC patients</p> Conclusion <p>The long-term survival advantage conferred by gastrectomy in patients with T1b GC was found to be commensurate with that achieved through endoscopic therapy. The prognostic model formulated in this study exhibited robust performance in forecasting the OS of T1b GC patients</p> Graphical abstract <p></p>