Background: Treatment of advanced non-small cell lung cancer with immune checkpoint inhibitors (ICI) has been shown to yield durable responses. High neutrophil-to-lymphocyte ratios (NLR), low lymphocyte-to-monocyte ratios (LMR) and high platelet-to-lymphocyte ratios (PLRs) have been used as surrogates for increased levels of inflammation in the tumor microenvironment that can predict cancer progression and response to therapy. However, the comparative prognostic role of these markers account for when immunotherapy is utilized has not been explicitly determined. Methods: We performed a retrospective review of 233 patients with advanced stage NSCLC at the University of Miami / Sylvester Comprehensive Cancer Center who received ICI either as first-line (1L) or second-line (2L) therapy and for whom laboratory data, in particular absolute neutrophil, lymphocyte, monocyte, and platelet counts, were available pre-treatment and 4 weeks after initiation of immunotherapy. Results: Using receiver operating characteristic (ROC) curves, we identified cutoff values with optimal prognostic value for NLR (4), LMR (2), and PLR (200). Median age, histology, and smoking history were equivalent across each group. Improved OS was observed in our cohort for patients in which pre-treatment NLR < 4 (54.4m vs. 35.9m, p = 0.0069), LMR > 2 (54.4m vs. 32.3m, p = 0.0016), and PLR < 200 (54.4m vs. 27.5m, p = 0.0007), while PFS was unaffected when looking strictly at these cutoffs. We then observed that PFS could be better predicted after stratifying NLR, LMR, or PLR when taking into account whether ICI was administered as 1L or 2L. We also determined that changes in (Δ) NLR or LMR (but not PLR) by at least 20% between baseline and 4 weeks after initiation of ICI could predict duration of response to ICI. Conclusions: In conclusion, the NLR, LMR, and PLR are powerful surrogates for the tumor microenvironment and can predict responses to ICI in advanced NSCLC when used in the context of previous lines of therapy and subsequent pro-inflammatory changes. Our study shows that ICI used in the first-line setting results in more durable responses, and can overcome an unfavorable tumor microenvironment. In addition, we demonstrate that in NSCLC, durability of responses can be predicted by changes in these systemic inflammatory response markers early after the initiation of ICI.