Background: Following radical prostatectomy, between 15 and 60% of all patients with pT3 prostate cancer experience persistence or increasing levels of prostate-specific antigen (PSA) as a sign of tumor persistence or progression within 5 years. Retrospective studies have shown a rate of 35–55% of positive biopsies from the vesicourethral anastomosis in this situation. Best treatment for these disease conditions is under debate, current strategies include adjuvant radiotherapy (RT), ‘wait-and-see‘ and salvage RT or hormone therapy for increasing PSA. Results: A number of retrospective studies have shown an increased rate of local control and ‘freedom from treatment failure‘ following adjuvant RT with doses in the range of 50–60 Gy. However, no survival benefit could be demonstrated by now. Results of three major phase III studies are pending. In case of persisting or increasing PSA levels following radical prostatectomy, 30–70% of these patients will reach an undetectable PSA level after conformal RT with total doses of 60–70 Gy, which will stay undetectable or at least stable within the next 2–5 years in about 50% and therefore offering a chance of cure. When starting RT, PSA should be as low as possible (<2 ng/ml). With higher PSA levels the chance of achieving an undetectable PSA again decreases below 35%. High Gleason scores of 8–10, seminal vesicle involvement and a short PSA doubling time are adverse prognostic factors. Severe late side effects of conformal RT are infrequent (<3%). In contrast, hormonal treatment is of palliative nature in the long run, with a median time to development of metastases of 4–7 years, and can be offered to patients with progressive disease after RT. Conclusions: Adjuvant RT following radical prostatectomy for pT3 prostate cancer offers higher local control rates and an increase in ‘freedom from treatment failure‘, but no prolongation of survival has yet been shown. In the situation of increasing PSA levels after radical prostatectomy, salvage RT seems to offer a chance of cure in selected patients, although it is difficult to draw firm conclusions because of generally too short follow-up times.