214 Background: FDA approval of three life-prolonging agents for metastatic castrate resistant prostate cancer (mCRPC) has occurred since April 2011. Two agents, abiraterone (abi) and enzalutamide (enza), disrupt androgen signaling. The third, radium-223 dichloride (Ra-223), targets bone metastases via alpha radiation. We sought to explore associations between prior abi/enza progression, number of Ra-223 cycles, prognostic factors, and overall survival (OS). Methods: Forty-two mCRPC patients (pts) treated with Ra-223 were identified. The sample was stratified based on progression (or not) on abi or enza prior to starting Ra-223. Number of Ra-223 doses administered, prognostic variables before Ra-223 treatment, and Kaplan-Meier estimates of overall survival (OS) were compared. Results: A strong association (p = 0.016) was demonstrated between prior abi/enza failure and number of Ra-223 doses administered. Patients without prior abi/enza failure were more likely to receive ≥ 4 doses of Ra-223 (94.12% vs 60.0%; odds ratio [OR] = 10.667; 95% CI, 1.214 – 93.699; p = 0.033). In comparison, those not receiving at least 4 doses of Ra-223 had a negative predictor in OS (p = .001) with a median survival of 89 days (n = 10; 95% CI, 44.064 – 133.936) compared to 303 days (n=15; 95% CI, 170.452 – 435.548). Patients initiating Ra-223 treatment after abi/enza progression had a higher median PSA (207.5 vs 49.2 ng/mL, p = 0.001), LDH (315 vs 253.5 U/L, p = 0.007) and alkaline phosphatase (ALP) (191 vs 106 U/L, p = 0.004). Conclusions: Our retrospective single institution analysis indicates that mCPRC patients who previously failed abi/enza are significantly less likely to complete 4 or more Ra-223 doses. These patients had significantly worse prostate cancer by a variety of standard prognostic variables. More information is needed from larger data sets to better understand this patient population and to best determine the optimal timing of Ra-223 administration.