e18680 Background: Abiraterone and enzalutamide are oral therapies widely used to treat men with castration-resistant prostate cancer (CRPC). Recent data have suggested potentially worsened quality of life and depression with use of enzalutamide compared to abiraterone. Because Veterans are at a higher risk for mental health conditions, we sought to compare mental health service utilization in Veterans with CRPC receiving enzalutamide to those receiving abiraterone. Methods: The Veterans Health Administration Corporate Data Warehouse was used to identify men with CRPC who received abiraterone or enzalutamide for ≥ 30 days as first-line treatment between 2010-2017. We compared the rate of mental health visits per 100 patient-months for men on abiraterone versus enzalutamide using an exact rate ratio test, assuming Poisson counts. Results: Among 2902 male Veterans, 68.6% (n=1992) received abiraterone and 31.4% (n=910) enzalutamide as first-line therapy. Men who received enzalutamide were older (76 vs 74, p<0.01) and had a higher comorbidity burden (Charlson Comorbidity Index [CCI] ≥ 2 in 28.7% vs 21.6%, p<0.01); no differences were noted in race or prevalence of preexisting documented mental health diagnoses. Median time on drug was 8 months for both medications. There was no difference in the rate of mental health visits per 100 patients-months on enzalutamide versus abiraterone (6.6 v. 6.7, p=0.66). However, within patient sub-groups, men who were age 75 or older, not married, or without notable comorbidities had lower rates of mental health visits with enzalutamide compared to abiraterone; whereas those who were younger than 75, married, had higher comorbidities, or a preexisting mental health diagnosis had higher rates of mental health visits with enzalutamide (Table). Conclusions: Among Veterans with CRPC who received a novel antiandrogen therapy first-line, there was no difference in engagement in mental health care between those who received abiraterone versus enzalutamide. Sub-group analysis revealed significant differences between patients on the two medications in demographic and diagnostic characteristics associated with number of visits, suggesting that vulnerability for mental health symptoms may vary by medication type. Further work in understanding the long-term impact of novel antiandrogens on mental health is needed.[Table: see text]