Background: Precision treatment of cancer uses biomarker-driven therapy to individualize and optimize patient care. Objective: To evaluate real-life clinical experience with biomarker-driven therapy in metastatic gastric and esophageal cancer in Israel. Patients and Methods: This multicenter retrospective cohort study included patients with metastatic gastric or esophageal cancer who were treated in the participating institutions and underwent biomarker-driven therapy. Treatment was considered to have a benefit if the ratio between the longest progression-free survival (PFS) post biomarker-driven therapy and the last PFS before the biomarker-driven therapy was ≥1.3. The null hypothesis was that ≤15% of patients gain such benefit. Results: The analysis included 46 patients (61% men; median age, 58 years; 57% with poorly-differentiated tumors). At least one actionable (i.e., predictive of response to a specific therapy) biomarker was identified for each patient. Immunohistochemistry was performed on all samples and identified 1–8 (median: 3) biomarkers per patient (most commonly: low TS, high TOPO1, high TOP2A). Twenty-eight patients received therapy after the biomarker analysis (1–4 lines). In the 1st line after biomarker analysis, five patients (18%) achieved a partial response and five (18%) stable disease; the median (range) PFS was 129 (12–1155) days. Twenty-four patients were evaluable for PFS ratio analysis; in seven (29.2%), the ratio was ≥1.3. In a one-sided exact binomial test vs. the null hypothesis, p = 0.019; therefore, the null hypothesis was rejected. Conclusions: Our findings demonstrated that implementing biomarker-driven analysis is feasible and could provide clinical benefit for a considerable proportion (~30%) of patients with metastatic gastric or esophageal cancer.[Figure not available: see fulltext.].