INTRODUCTION: There is substantial variability in survival and response outcomes for patients using immune checkpoint inhibitors (ICIs), and predictive markers are required to guide treatment decisions. A lung immune prognostic index (LIPI) was recently developed to predict ICI treatment outcomes and we aim to evaluate whether LIPI is predictive of survival and response for patients treated with atezolizumab for advanced NSCLC. METHODS: Pooled analysis was performed of patient-level data of four clinical trials of atezolizumab for NSCLC. Based on pre-treatment derived neutrophil-to-lymphocyte ratio greater than 3 and lactate dehydrogenase greater than the upper limit of normal, patients were assigned to one of three groups: good LIPI, 0 risk factors; intermediate LIPI, 1 risk factor; or poor LIPI, 2 risk factors. The primary outcome was overall survival (OS). RESULTS: In a pooled cohort of 1,489 patients treated with atezolizumab, the LIPI group was significantly associated with OS (p < 0.001), progression-free survival (p < 0.001), and response (p < 0.001). Median OS for good (n = 678), intermediate (n = 631), and poor (n = 180) LIPI groups was 18.4 months, 11.3 months, and 4.5 months, respectively. Association between LIPI and OS was consistently identified across a range of atezolizumab-treated subgroups. For docetaxel-treated patients (n = 687), LIPI was similarly associated with survival (p < 0.001) and response (p = 0.005). CONCLUSIONS: Pre-treatment LIPI is a convenient prognostic marker able to identify atezolizumab-treated patient groups with significantly different survival and response outcomes. However, LIPI is also a prognostic marker of survival and response for patients treated with chemotherapy; thus, it is not specifically predictive for ICI treatment.