We conducted a meta-analysis of randomized trials comparing regimens that included daily oral prednisone (P) in only one arm to investigate its impact on toxicities and outcomes in metastatic castration-resistant prostate cancer (mCRPC). Five trials were identified totaling 2939 patients, of whom 1471 were randomized to an arm not containing P and 1468 received therapy containing P. There was no difference between the non-P and P groups for severe toxicities (incidence rate ratio [IRR]=0.82, p=0.712, I2=97.9%). When examining toxicities as a reason for discontinuing therapy, the non-P groups were not different from the P groups (relative risk [RR]=1.24, p=0.413, I2=86.8%). The non-P groups demonstrated no difference in OS compared to the P groups (HR=1.09, p=0.531, I2=79.7%). The meta-analysis is limited by the trial level design and small number of trials. © 2013 Elsevier Ireland Ltd.