Objective: To investigate the prognostic predictive value of a nomogram constructed based on albumin-related inflammatory composite indicators for patients with advanced non-small cell lung cancer (NSCLC) treated with immune checkpoint inhibitors (ICIs). Methods: The clinical data of 274 patients with advanced NSCLC were retrospectively analyzed. The overall survival (OS) was set as the primary endpoint. Independent prognostic factors were selected through the Cox proportional hazards model and a nomogram was constructed. The performance of the model was evaluated using C-index, time-dependent ROC curve, calibration curve, and DCA curve, etc. At the same time, the prognostic risk stratification ability of the model was analyzed. Results: Liver metastasis, TNM stage IV, CEA ≥ 2.54 ng/mL, and high-risk group of NPS were independent risk factors for OS. CALLY ≥ 0.6 was an independent protective factor. The AUC of the nomogram for 1-year and 3-year OS was 0.755 and 0.801, respectively, and the C-index was 0.72. The calibration curve and DCA confirmed the good fit of the model and clinical net benefit, and the patients could be stratified into low, medium, and high-risk groups. The stratification ability was significant (P<0.01). Conclusion: The nomogram constructed based on liver metastasis, TNM stage, CEA, NPS, and CALLY can effectively predict the OS of patients, and the prognostic risk stratification ability is good. The model has excellent discrimination and calibration, and has high clinical practical value, which can provide precise references for the formulation of individualized immunotherapy strategies.