Background: Immune checkpoint inhibition (ICI) has improved outcomes for many treatment-naïve advanced non-small cell lung cancer (NSCLC) patients. However, better biomarkers are needed to predict patient response and guide treatment decisions considering added toxicity and higher cost of combination treatments. A prospectively designed, observational study assessed the ability of a clinically validated, blood-based, host immune classifier (HIC) to predict ICI therapy outcomes. Methods: The study (NCT03289780) includes 33 US sites having enrolled over 3,000 NSCLC patients at any stage and line of therapy. All enrolled patients are tested and designated HIC-Hot (HIC-H) or HIC-Cold (HIC-C) prior to therapy initiation. An interim analysis of secondary and exploratory endpoints was performed after 12- 18 months (mo) follow-up with the first 2,000 enrolled patients. We report the overall survival (OS) of HIC-defined subgroups comprising advanced stage (IIIB and higher) NSCLC patients treated with first-line regimens (284 ICI containing treatments, 877 total first-line patients). Results: In a real-world clinical setting, OS of advanced stage NSCLC treatment-naïve patients receiving platinum-based chemotherapy (n = 392) did not differ significantly from patients receiving any type of ICI containing regimen (n = 284); 11.7 mo vs. 14.4 mo; hazard ratio (HR) = 0.94 [95% confidence interval (CI): 0.76–1.17], p = 0.59. HIC-H patients experienced longer survival than HIC-C across multiple regimens, including ICI. For all ICI, median OS (mOS) was not reached for HIC-H (n = 196, CI: 15.4 mo–undefined) vs. 5.0 mo (n = 88, CI: 2.9 mo–6.4 mo) for HIC-C patients (HR = 0.38 [CI: 0.27–0.53], p < 0.0001). Similar results were seen in the ICI only (16.8 mo vs. 2.8 mo; n = 117, HR = 0.36 [CI: 0.22–0.58], p < 0.0001) and ICI/chemotherapy combination subgroups (unreached vs. 6.4 mo; n = 161, HR = 0.41 [CI: 0.26–0.67], p = 0.0003). In the PD-L1 high cohort (PD-L1 ≥50%), mOS for HIC-H was not reached (n = 81, CI: 13.9 mo–undefined) vs. 3.9 mo (n = 41, CI: 2.1 mo–7.8 mo) for HIC-C (HR: 0.39 [CI: 0.24-0.66], p = 0.0003). HIC results were independent of PD-L1 score (p = 0.81) and remained predictive of OS in first-line ICI-treated patients when adjusted for PD-L1 and other covariates by multivariate analysis (HR = 0.40 [CI: 0.28-0.58], p < 0.0001). Conclusions: Blood-based host immune profiling may provide clinically meaningful information for selecting NSCLC patients for two common ICI containing regimens independent of and complementary to PD-L1 score.