Purpose: To identify pre-therapeutic variables associated with overall survival (OS) in patients treated with 223Ra. Methods: Data from 45 CRPC patients treated with 223Ra were retrospectively analyzed. All patients who received at least one 223Ra injection were included in the study. Cox proportional hazard regression models were used to estimate hazard ratio's (HR) and to test for association. Results: Twenty-one patients (47%) received six 223Ra injections and 24 patients (53%) received one to five 223Ra injections. Median OS since start of 223Ra was 13.0 months (95% confidence interval (CI) 8.2-17.8). Patients who completed 223Ra therapy had a median OS of 19.7 months (95% CI 14.9-24.6), while patients who received one to five 223Ra injections had a median OS of 5.9 months (95% CI 3.8-8.1; P < 0.001). Univariable analysis showed poor baseline ECOG performance status (PS), baseline opioid use, lowered baseline hemoglobin, and elevated prostate-specific antigen, alkaline phosphatase and lactate dehydrogenase (LD) levels were significantly associated with OS. Multivariable Cox regression analysis demonstrated that poor baseline ECOG PS (HR 10.6) and high LD levels (HR 7.7) were pre-therapeutic variables that predicted poor OS. Conclusions: In a multivariable Cox regression model, good baseline ECOG PS and low LD levels were significantly associated with longer OS in patients treated with 223Ra. These variables may be used for stratification of CRPC patients for 223Ra therapy. Prospective studies to evaluate these variables are warranted, to develop a nomogram to select patients properly. In this retrospective study, predictors of overall survival in 45 metastatic castration-resistant prostate cancer patients treated with 223Ra therapy were evaluated. Baseline ECOG performance status and lactate dehydrogenase levels turned out to be significant in a multivariable prediction model for overall survival.