Background: The optimal treatment (tx) of men with castration sensitive prostate cancer (CSPC) who biochemically recur following a radical prostatectomy (RRP) is not known. Data from E3805 suggest that men with metastatic CSPC live longer if they receive D in addition to ADT relative to ADT alone. We tested this hypothesis in non-metastatic CSPC, via a phase III study (TAX3503). Methods: PC pts who underwent an RRP with BCR and a doubling time ≤ 9 months (mo) were eligible. PSA ≥ 1 ng/mL and testosterone ≥100 ng/dl were required. Randomization was 1:1 to receive leuprolide 22.5 mg q3 mo x 18 mo, bicalutamide 50 mg x 30 days, with D at 75 mg/m2 q3 weeks x 10 cycles (Arm A) or without D (Arm B). The primary endpoint was PFS defined as a detectable PSA or death. Pts with T recovery >50 ng/dl were evaluable. The intent to treat population (ITT) was also examined. A sample size of 412, to yield 370 evaluable pts, was calculated to detect a HR of 1.6 with 90% power. The trial was closed by the sponsor after the pts completed treatment; remaining pts were then followed to PFS via a registry for 18 mo’s. A test to detect a difference in the hazard rates between arms was generated by the log rank statistic. Results: 413 pts were randomized. Median f/u time in the ITT population was 31.5 mo’s (0.0-60.2). Data are summarized below in the Table. Conclusions: The clinical benefit of ADT + D relative to ADT alone in men with high-risk BCR after RRP appears to be marginal, although there is a statistical trend towards improved PFS. Data are limited by the biases intrinsic to post-protocol registries and by short duration of follow up, although tracking of pts continues. Clinical trial information: NCT01813370