From 1966 to 1979, 360 patients with clinical stages A2, B and C1 prostate cancer underwent staging pelvic lymphadenectomy, and completed a course of combined interstitial radioactive gold seeds and external beam radiotherapy. All patients had a normal serum prostatic acid phosphatase level and a bone scan negative for metastases. All patients were followed until death or for a mean of 7.3 years (range 1.2 to 18.25 years) for those alive at analysis. To determine the risk of dying of prostate cancer we reviewed the records of the 142 patients (39%) who died. At analysis 21% of the patients had died of prostate cancer and 17% of other known causes. The cause of death could not be determined in 4 patients (1%). Cardiovascular disease accounted for a fifth of all deaths. The actuarial risk of death of prostate cancer for all patients was 8 +/- 3% (+/- 2 standard errors) at 5 years and 30 +/- 7% at 10 years. The risk of death of all causes was 16 +/- 4% at 5 years and 46 +/- 7% at 10 years. An increased risk of cancer death was associated with established risk factors, including advanced local disease, poorly differentiated histology, pelvic nodal metastases and distant recurrence. We also noted a substantial risk of cancer death in patients who had local tumor recurrence. While previous studies have reported a relatively low incidence of cancer deaths (4 to 17%) in patients initially diagnosed with localized disease, our data suggest that prostate cancer is the major cause of mortality in such patients. Aggressive curative therapy, regardless of treatment modality, should be considered for localized prostate cancer in men with a life expectancy of 10 or more years.