Background: The neutrophil–lymphocyte ratio (NLR) is recognized as a prognostic marker in many cancers, including metastatic castration-resistant prostate cancer (mCRPC). In this study, we perform a retrospective analysis of the pivotal COU302 study of abiraterone acetate (AA) as first line therapy for men mCRPC. Methods: The COU302 study randomized men with minimally symptomatic mCRPC to 1000mg AA once daily plus 5mg prednisone (AA + P) twice daily or placebo plus 5mg prednisone (P) twice daily. We utilized Kaplan–Meier survival and cox proportional hazard analyses to assess the effect of baseline NLR on response to AA + P versus P. Adjustment co-variates were selected from a recently reported analysis of the PREVAIL study of enzalutamide in men with mCRPC. Based on this analysis, we selected a NLR cut-off value of 2.5. Outcomes assessed included the co-primary endpoints from the COU302 study overall survival (OS) , radiographic progression free-survival (RPFS) and PSA progression-free survival (PFS). Results: Among the 1088 patients in the COU302 study, baseline NLR values showed significant differences according to baseline albumin below median (3.16 vs 2.92, p = 0.011), but no other baseline covariates. Mean NLR significantly increased following treatment initiation and at the end of study compared to baseline. Among patients with a baseline NLR ≥2.5, there was similar OS in AA + P and P treatment arms (p = 0.26), which was confirmed on adjusted cox proportional hazards (AA +P HR 0.95, p = 0.65). AA + P showed a significant OS benefit versus P (p = 0.0011) in men with baseline NLR <2.5, with an adjusted HR of 0.723(p = 0.012). For RPFS, benefits for AA + P were seen in men with NLR <2.5 or ≥ 2.5, with the magnitude of benefit greater in men with NLR <2.5. In the AA + P arm, men with baseline NLR <2.5 had significantly better PSA PFS compared to baseline NLR ≥2.5 (p = 0.03); no significant differences were seen in men in the placebo arm. Conclusions: In the COU302 study, we observed that patients with a screening NLR<2.5 had a significant benefit to AA + P compared to P, whereas for patients with a screening NLR ≥2.5, this benefit was less evident. Increases in the NLR value from baseline could not be used to predict response to AA. Legal entity responsible for the study: The authors. Funding: Has not received any funding. Disclosure: V. Fradet: Speaker: Abbvie, Bayer, Ferring, Janssen, Sanofi; Consultant: Amgen, AstraZeneca, Astellas, Bayer, Ferring, Janssen, Sanofi; Adviser: Amgen, Astellas, Bayer, Ferring, Janssen, Sanofi; Research grant: Amgen, AstraZeneca, Astellas, Janssen, Sanofi; Clinical trial: AstraZeneca, Bayer, Janssen; Meeting sponsor: Bayer, Janssen, Lilly; Educational grant: Janssen. P. Toren: Scientific study or trial: Innocrin Pharma, Roche; Consultant, Advisor: Sanofi Canada, Ferring, Astellas, Abbvie. All other authors have declared no conflicts of interest.