Background: Relatively limited clinical data exists for patients who receive T-DXd treatment in the 3L+ setting for HER2-positive unresectable advanced/recurrent gastric or gastroesophageal junction cancer. Insufficient data exists on real world patient populations that are likely to be excluded from clinical trials (e.g., elderly patients, those with significant co-morbidities). Therefore, this study was designed to evaluate the real-world clinical effectiveness and safety of T-DXd in patients with HER2-positive gastric cancer. Methods: This study (UMIN000049032) was a real-world, multicenter, non-interventional, retrospective cohort study conducted in Japan. Patients with HER2-positive unresectable advanced/recurrent gastric or gastroesophageal junction cancer treated with T-DXd were enrolled. The patients were first administered T-DXd between Sep 2020, when T-DXd became available for clinical use in Japan, and Sep 2021. We followed up on these patients' records until Sep 2022. The endpoints were overall survival (OS), real-world progression-free survival (rwPFS), time to treatment failure (TTF), objective response (ORR), adverse events (AEs) of Grade ≥3, and adverse events leading to dose reduction, interruption, or discontinuation. Results: A total of 318 patients were enrolled, and 312 patients were analyzed, excluding 6 patients due to double enrollment or miss-enrollment (median age 70 years, performance status 0 or 1/≥2 [87%/12%], 43% patients with ascites). Median OS, rwPFS, and TTF were 8.90 months, 4.57 months, and 3.94 months, respectively. Overall ORR was 32.7%; it was 43.8% in the 226 patients with a target lesion. Regarding safety, the incidence of Grade ≥ 3 AEs was 48.4%. A total of 190/312 (60.9%) patients treated with T-DXd experienced dose reduction, interruption, or discontinuation of T-DXd treatment due to AEs. The most common reasons for dose reduction or interruption were neutrophil count decreased, anorexia, and malaise, while the most common reasons for discontinuation were interstitial pneumonia, anorexia, malaise, and nausea. Conclusions: This study demonstrates the effectiveness and safety of T-DXd in the 3L+ setting in real-world patients with HER2-positive gastric or gastroesophageal cancer. No new safety signals were identified on the basis of this study.