Purpose <p>Neutrophil-to-lymphocyte ratio (NLR) is a widely studied variable to investigate the systemic inflammation-related prognostic effect, and in some clinical setting has been able to predict the activity of immunotherapy. Many studies confirmed a negative relationship of pre-treatment high NLR and the outcome of patients with esophageal squamous cell cancer (ESCC), but poor data are available for patients with unresectable or metastatic ESCC (mESCC). The study aims to perform a trial-level analysis of the relationship of NLR with overall survival (OS) among mESCC patients receiving an upfront systemic treatment.</p> Methods <p>A systematic literature review of studies enrolling mESCC patients receiving an upfront chemotherapy-based regimen and reporting the hazard ratio and confidence intervals of the relationship of NLR with OS were carried out. A meta-analysis was performed, calculating a global effect size, and another one evaluated only immunotherapy-based studies. Finally, based on the resulting between-study heterogeneity, a meta-regression evaluated the possible inference of the more frequently reported baseline variables on the relationship.</p> Results <p>Twenty-three articles, including 24 study cohorts, were selected, including 3590 patients. The meta-analysis calculated a significant effect size for the relationship of NLR with OS (HR 2.01, CI 1.64–2.45), that was maintained among the 16 immunotherapy-based studies (HR 2.12, CI 1.58–2.84). The significant heterogeneity between studies (<i>Q</i>&#xa0;=&#xa0;126.24, <i>p</i>-value&#xa0;&lt;&#xa0;0.0001; <i>I</i>²&#xa0;=&#xa0;81.8%) allowed to perform the planned meta-regressions. However, none of the 10 baseline variables was able to influence significantly the relationship of NLR wih OS.</p> Conclusions <p>Baseline NLR is a reliable prognostic factor of mESCC patients receiving upfront chemotherapy-based regimens, with a lightly more pronounced effect size when the regimen included.</p>