The appropriate management of elderly patients diagnosed with prostate cancer remains controversial. In order to provide guidelines as to when aggressive local treatment may be indicated, we provide estimates of the long-term probability of death from prostate cancer and other competing causes in patients of 70 years of age or older, who underwent radical surgery in the form of radical perineal prostatectomy for clinically non-metastatic disease. In this study, a total of 484 patients with an age of 70 or above who underwent radical perineal prostatectomy between 1970 and 2000 comprised a retrospective cohort of patients with clinically organ confined prostate cancer. Of these patients, 461 patients (95.3%) had a minimum follow-up of half a year and were included in the analysis. The median age was 73 years (range 70-81 years) and the median follow-up was 5.4 years. Overall 115 men died during the follow-up period with 49.6% of deaths attributable to prostate cancer. The median time to cancer-associated death was 17.5 years and the median time to death of any cause 11.6 years. When the likelihood of death from prostate cancer as a function of Gleason score was estimated, the 10-year cancer-associated death rates of patients with Gleason scores of 2-6, 7 and 8-10, were 15.2, 25.2 and 40.2%, respectively. In the subset of patient with margin positive disease the estimated likelihood of a cancer-associated death was 45.3% after 10 years. While the median time to cancer-associated death for margin positive patients with a Gleason score of 2-6 was not reached, patients with a Gleason score of 7 and 8-10 experienced median cancer-associated survival times of 9.6 and 7.6 years, respectively. In conclusion, Gleason score is a strong predictor of the likelihood of prostate cancer related death in elderly patients. Patients with a given Gleason score and a projected life expectancy of at least 10 years may be at similar risk of dying of prostate cancer as younger patients.