Background: There is emerging interest in the use of bone turnover markers in CRPC to compliment current modalities and aid therapeutic management decisions. We report the preliminary results of investigation into the clinical utility of P1NP, total alkaline phosphatase (ALP) and PSA as parameters of survival in a representative sample of patients from TRAPEZE. Methods: Baseline and post treatment (cycle six or end of treatment if sooner) samples were taken from 120 patients equally distributed across the trial arms of TRAPEZE. Absolute and percentage change in the markers were investigated in relation to overall survival (OS) using the Cox Regression model. Results: Of the 120 patients, 38(32%) failed to complete 6 cycles. No relationship was observed between P1NP and OS using either absolute or percentage change (p=0.59, p=0.64). Absolute change in log PSA was significantly related to OS (HR=1.22 (95%CI 1.01, 1.48), p<0.05, respectively) however percentage change in log PSA did not reach statistical significance (HR=1.006 (95%CI 0.99, 1.01), p=0.095). Absolute change in ALP was found to be inversely related to OS (HR=0.999 (95%CI 0.998, 0.999), p<0.001) but percentage change in ALP was not significant (p=0.89). When the PSA and ALP models were applied to the entire trial set (757 patients) the predictive value of both absolute and percentage change in log PSA remained highly significant (HR=1.23, p<0.001 & HR=1.01, p<0.001). In addition the significance of absolute change in ALP also remained (HR=0.999, p<0.05). Conclusions: Changes in P1NP were not related to OS in CRPC patients metastatic to bone. Absolute change in ALP was inversely related to OS and both absolute and percentage changes in log PSA were also related to OS in this sub-study. When applied to the entire TRAPEZE patient group all remained statistically significant. The utility of changes in ALP and PSA, as prospective markers for studies in CRPC patients metastatic to bone, merit further evaluation.