Purpose: To compare a standard radio-oncological and a surgical biochemical failure definition after combined-modality radiation therapy (CRT) in men with intermediate- and high-risk prostate cancer. Methods: 425 men were treated with external beam radiotherapy (59.4 Gy, 33 fractions) and 125J seed-brachytherapy (S-BT, 100 Gy). Biochemical recurrence (BR) was defined either as radio-oncologic (rBR), using a +2 ng/mL prostate-specific antigen (PSA) increase above a nadir value, or as surgical (sBR), using a 2-year posttreatment PSA of ≥0.2 ng/mL. Biochemical recurrence-free, metastasis-free, cancer-specific, and overall survival were calculated at 5 and 10 years using the Kaplan-Meier method. Standard validation tests were used to compare both thresholds. Results: After a median of 7 years, overall recurrence rates were 10.4% and 31.5% for rBR and sBR definitions, respectively. Both failure definitions proved sensitive for the prediction of metastases and cancer-specific death, whereas the rBR definition was significantly more specific. The accuracies of a correct prediction of metastases and death of prostate cancer were 73.1% vs. 96.2% and 72.2% vs. 92.9% for sBR vs. rBR, respectively. The inferior validity results of the sBR definition were attributable to a PSA-bounce phenomenon occurring in 56% of patients with sBR. Still, using the less suitable sBR definition, the results of CRT compared favorably to BRFS rates of surgical interventions. Conclusion: After CRT, the radio-oncological (aka Phoenix) failure definition is more reliable than a fixed surgical endpoint. Exclusively in high-risk patients, sBR offers a direct comparison across surgical and nonsurgical treatment options at 5 and 10 years.