Background: The objective of this study was to evaluate real-world OS in men with chemotherapy-naïve mCRPC treated with first-line ENZA or ABI or sequencing to these agents after treatment with BIC ( > 90 d). Methods: A retrospective analysis was performed using the Veterans Health Administration (VA) database. Men with mCRPC and ≥ 1 pharmacy claim for ABI or ENZA (1st claim date = index date) following surgical or medical castration and with no chemotherapy treatment 12 months pre-index date were identified from 01APR2014 to 31MAR2017. Men had continuous VA enrollment for ≥ 12 months pre- index date and were followed until death or study end. Cox proportional hazards regression models examined the impact of treatment on OS, and the impact of first using BIC ( > 90 d) vs first-line use of ABI and ENZA on OS. Adjusted Kaplan-Meier analysis was conducted to graphically describe OS. Results: This study included 1229 ENZA- and 1945 ABI-treated men with mCRPC with mean ages of 74 and 73 y, respectively. Median follow‐up was 548 and 572 d, and median OS was 892 and 786 d, respectively. ENZA and ABI were first-line treatment in 1068 and 1628 men, but BIC ( > 90 d) was first used before sequencing to ENZA or ABI in 161 and 317 men, respectively. Median duration of BIC treatment was similar in men subsequently treated with ENZA or ABI (251 v 246 d, respectively). Compared with first-line ABI, use of ABI following BIC led to shorter OS (hazard ratio [HR] = 1.30; 95% CI 1.11 -1.52). Compared with first-line ENZA, use of ENZA following BIC resulted in a nonsignificant effect on OS (HR = 0.92; 95% CI 0.72-1.19). In the Cox analysis, ENZA-treated men had longer OS compared with ABI-treated men (HR = 0.84; 95% CI 0.76-0.94). Conclusions: Use of BIC ( > 90 d) before sequencing to ENZA did not negatively effect OS, however BIC before ABI impacted OS adversely in men with chemotherapy-naïve mCRPC. There are significant differences in OS favoring ENZA compared with ABI regardless of prior BIC. Real-world observational data are nonrandomized, and markers for disease severity/volume are not available in the claims database. Further research is required to confirm these findings.