Background: Androgen deprivation therapy (ADT) served as standard of care for mCSPC for decades. Combining ADT with docetaxel (DOC) starting in 2015 or with novel hormonal therapies (NHT) starting in 2017 has demonstrated improved survival compared to ADT alone. This study examined the impact of new evidence on treatment selection for mCSPC patients in real-world oncology practice. Methods: Electronic medical record (EMR) data from a network of U.S.-based oncology practices in the ConcertAI Oncology Dataset were used to retrospectively evaluate treatment patterns in mCSPC patients who initiated first-line (1L) therapy between 2014–2019. We estimated the proportion of mCSPC patients receiving each 1L regimen, duration of therapy until initiation of next regimen, and trends in 1L regimen. Treatment patterns were studied in the overall population and by race/ethnicity. Anti-androgen (AA) use for <90 days was not included. Results: A total of 858 mCSPC patients were included (70% White, 16% Black, 3% Hispanic, and 11% other race/unknown). Median age at mCSPC diagnosis was 69 years, and 63% presented with de novo metastases. The most common mCSPC 1L regimens were ADT + older AA (26.3%, mainly bicalutamide), ADT monotherapy (20.5%), ADT + NHT ± AA (19.2%), and ADT + DOC ± AA (16.4%). The remaining 17.5% received a variety of other therapies, including AA monotherapy (5.9%) or NHT ± AA (5.5%). NHT included abiraterone, apalutamide, and enzalutamide. ADT + NHT ± AA treatment increased each year, while ADT + DOC ± AA treatment peaked in 2017 and then decreased (Table). By contrast, ADT + AA was the most common therapy in 2014 but declined every year. Median duration until initiation of a subsequent regimen was 14.3 months for ADT + NHT ± AA and 10.8 months for ADT + DOC ± AA. Differences in 1L treatment patterns across White and Black patients were not statistically significant in unadjusted analyses. Conclusions: Even in 2019, over half of mCSPC patients treated in real-world settings did not receive 1L therapy now known to significantly improve survival (ADT + NHT or ADT + DOC) over ADT alone. Those who did, received shorter durations of treatment than observed in registrational trials. However, we found no initial evidence of racial disparities in treatment. The disconnect between trial evidence and real-world practice could be due to patient/disease characteristics, cost/access issues, or provider awareness. A better understanding of these contributing factors is worthy of further study.