Although prostate cancer is currently the second leading cause of cancer deaths in men, the treatment of localized prostate cancer remains controversial. The fundamental problem is a lack of knowledge concerning the natural history of the disease and the absence of adequate follow up data concerning the efficacy of current treatment strategies. Three factors have created a new urgency to determining the appropriate treatment for localized prostate cancer: 1. American men are surviving well into their 70's and 80's when asymptomatic prostate cancer is widely prevalent, 2. new technology including rectal ultrasound and the spring loaded biopsy gun has been adopted which permits the identification of previously undetectable cancers, and 3. urologists have shown a renewed enthusiasm for aggressive treatment of early prostate cancer. Treatment costs are well over $10,000/case, but do they benefit the older male? The appropriate therapeutic choice revolves around the question: For men with newly diagnosed prostate cancer, what increase in life expectancy and decreased morbidity is gained by the morbidity and expense of aggressive therapy? To address this question, a cohort of 1463 men diagnosed with localized disease between 1971 and 1980 has been identified using the Connecticut State tumor registry. A minimum ten year follow up is available on 89% of the cases. Survival analysis of this cohort suggests a constant 2% annual excess mortality rate due to prostate cancer. This project proposes a detailed retrospective analysis of this patient cohort in order to calculate the disease specific mortality rate for untreated adenocarcinoma of the prostate. This rate will be compared with the disease specific mortality rate associated with patients undergoing radical surgery, radiation therapy and immediate endocrine manipulation. Patient cohorts will be controlled for age, tumor histology and patient co- morbidities within the limits of a retrospective study. The information provided will yield new insights into the natural progression of prostate cancer and provide the necessary statistical and ethical foundation for constructing case control studies and treatment algorithms.