Background: Abiraterone (AA) and enzalutamide (ENZ) are two second generation antiandrogens used to treat advanced prostate cancer, but no large head-to-head trials have been performed. These oral therapies are commonly used in older patients with medical comorbidities who are not candidates for chemotherapy or clinical trials and have different mechanisms of action, adverse events, and drug interactions. To understand survival of patients with prostate cancer, we studied United States veterans treated prior to approval of AA and ENZ for metastatic hormone sensitive prostate cancer when both drugs had approval for metastatic castration resistant prostate cancer. Methods: We identified patients treated with AA or ENZ between 9/10/2014 and 6/3/2017 in the Veterans Health Administration and followed them to April 2020. Age, Elixhauser comorbidity score, treatment with androgen deprivation therapy (ADT) and docetaxel were collected. Cox proportional hazards modeling was used to assess the association between first oral treatment (AA or ENZ) and overall survival, while adjusting for covariates. Results: Of 5895 patients, 2562 (43.5%) were initially treated with ENZ, 3333 (56.5%) with AA, and 3040 (51.6%) received only one of the two drugs during the study period. Patients initially treated with ENZ compared to AA were older (mean 75.9 vs. 75.0 years, p = 0.001), had higher mean comorbidity score (6.2 vs. 5.9, p < 0.001), and were less likely to receive both ENZ and AA (45.2% vs. 51.0%, p < 0.001) or docetaxel (24.1% vs. 28.4%, p < 0.001). Patients who received only AA or ENZ and never received docetaxel were older (mean 78.3 vs. 73.2 years, p < 0.001) with higher mean comorbidity scores (6.4 vs. 5.7, p < 0.001). In the entire cohort, initial treatment with ENZ was associated with longer median survival (24.1 vs. 22.2 months, p = 0.003). After adjusting for age and comorbidities, ENZ was associated with a decreased risk of death compared to AA (HR 0.87, 95% CI 0.82-0.92). In 3317 patients who received two or more therapies (ENZ, AA, docetaxel) there was no difference in median survival between initial treatment with ENZ or AA (28.0 vs. 27.9 months). In 2578 patients (43.7%) who never received docetaxel and either ENZ or AA only, median survival was longer in patients treated with ENZ (18.9 vs. 13.6 months, p < 0.001) and was associated with decreased mortality when adjusting for age and comorbidities (HR 0.73, 95% CI 0.67-0.80). Conclusions: In the overall cohort, initial treatment with ENZ was associated with increased survival compared to AA. Patients who received only ENZ or AA and never received docetaxel had the largest benefit from ENZ, a difference of 5.3 months median survival. Efforts should be made to improve therapy selection for patients with prostate cancer, especially older patients with comorbidities.