ObjectiveOur study represents the first meta-analysis conducted to evaluate the prognostic utility of the baseline prognostic nutritional index (PNI) in patients with gastrointestinal cancer (GIC) who received immune checkpoint inhibitor (ICI) therapy.MethodsWe searched PubMed, the Cochrane Library, EMBASE, and Google Scholar until April 23, 2023, to obtain relevant articles for this study. Our analysis examined several clinical outcomes, including overall survival (OS), progression-free survival (PFS), objective response rate (ORR), and disease control rate (DCR).ResultsIn this analysis, a total of 17 articles with 2883 patients were included. Our pooled results indicated that patients with high PNI levels had longer OS (HR: 0.530, 95% CI: 0.456-0.616, p < 0.001) and PFS (HR: 0.740, 95% CI: 0.649-0.844, p < 0.001), as well as higher ORR (OR: 1.622, 95% CI: 1.251-2.103, p < 0.004) and DCR (OR: 1.846, 95% CI: 1.428-2.388, p < 0.001). Subgroup analysis showed that PNI cutoff values of 40 to 45 showed greater predictive potential. Subgroup analysis also confirmed that the above findings still hold true in patients with esophageal cancer, gastric cancer, and hepatocellular carcinomas.ConclusionThe PNI were reliable predictors of outcomes in GIC patients treated with ICIs.