To define the optimal role for radiotherapy (RT) after radical prostatectomy (RP) and to characterize specific patterns of PSA failure in this setting. The records of 105 patients who underwent RT after RP (69 received therapeutic RT because of an elevated prostate-specific antigen [PSA] level, 36 received immediate adjuvant RT) were reviewed. The median follow-up was 35 months after RT and 57 months after RP. Radiation success was defined as achievement and maintenance of a PSA less than 0.2 ng/mL. Preoperative, pathologic, and postoperative characteristics were examined for their ability to predict success after RT. Patterns of PSA recurrence after RT were also examined by determining the PSA nadir, PSA velocity, and timing of androgen-deprivation therapy. Of 105 patients, 47 experienced biochemical failure. Actuarial 3 and 5-year progression-free survival estimates for all patients were 55% and 43%, respectively. Significant favorable predictors of response to RT by multivariate analysis were preoperative PSA less than 20 ng/mL and the use of adjuvant RT. However, patients who received therapeutic RT with a pre-RT PSA less than 1.0 ng/mL demonstrated progression-free outcome equivalent to those who received adjuvant RT. Two distinct patterns of PSA failure were observed on the basis of PSA nadir after RT. Patients whose PSA failed to reach a nadir less than 0.2 ng/mL after RT had progression with a high PSA velocity (1.5 ng/mL/yr). Patients whose PSA reached a nadir less than 0.2 ng/mL but who subsequently had treatment failure progressed later with a lower PSA velocity (0.36 ng/ml/yr). RT is effective in select patients after RP. Given the low PSA velocity consistent with persistent local disease in nearly 50% of patients in whom RT failed, more effective local therapy is needed after RP in high-risk patients.