Background: Immune checkpoint inhibitors (ICIs) have been frequently used in cancer therapy and found to result in immune-related adverse effects due to nonspecific immune activation. Nanoparticle-based delivery systems have been proposed as a strategy to improve tumor targeting and reduce systemic toxicity; however, their safety and immunotoxicity profiles have not been comprehensively synthesized. Method: This study adopted a systematic review design following PRISMA guidelines. The two databases of PubMed and Scopus were searched for articles published from 2015 onward concerning the application of nanoparticle-mediated delivery of PD-1, PD-L1, or CTLA-4 inhibitors in solid tumors. Inclusion criteria were applied for studies on preclinical and early-phase clinical studies reporting safety, toxicity, or immune-related outcomes. Data on nanoparticle platforms, safety parameters, and methodological quality were extracted and bias risks were assessed against the SYRCLE tool. Results: Among the twenty-four studies meeting the inclusion criteria, the majority reported preclinical animal studies. Polymeric nanoparticles were the most commonly investigated platforms, followed by lipid-based, inorganic, albumin-based, and gene-delivery systems. Most studies reported stable body weight, preserved organ histology, and largely normal serum biochemical profiles. Cytokine analyses most commonly reported changes in IFN-γ, TNF-α, and, less frequently, IL-6, suggesting that nanoparticle-mediated ICI delivery may induce immune activation while only limited evidence of acute systemic inflammatory toxicity was observed in the included short-term studies. Risk of bias assessment revealed adequate randomization and outcome reporting but frequent uncertainty in blinding and housing procedures. Conclusion: Nanoparticle-mediated therapies demonstrate encouraging short-term safety profiles. However, long-term immunotoxicity, nanoparticle persistence, and clinical translation remained insufficiently addressed, suggesting the need for standardized safety reporting and expanded clinical evaluation.