Abstract Background Failure after radical prostatectomy can occur even out to 25 years after surgery. Therefore, it is important that studies have sufficient follow-up to determine more accurately the risk of failure. We evaluated a large cohort of patients for pathologic findings and risk of failure with a median follow-up of 9.5 years. Material and Methods Between 1985 and 1995, 719 patients underwent radical prostatectomy for lymph node negative prostate cancer. The prostate was inked and evaluated for: (1) positive bladder neck or urethral margin, (2) positive seminal vesicle, (3) into capsule, (4) through capsule, and (5) positive margin. These were considered positive pathologic findings. Results Overall, 264 (37%) of the patients had biochemical recurrence. For those patients with failure, median time to biochemical recurrence was 2.4 years. Five and 10-year biochemical failure rates were 28% and 38%, respectively. Pathologic stage of disease significantly (<0.0001) predicted for subsequent failure. If there were no positive pathology findings, the recurrence rate was 25%, compared to 63% for any of the 3 findings. Overall, 212 (29%) of the patients have died. Five and 10-year survival were 91% and 75%, respectively. A total of 45 patients (6%) died of prostate cancer. For patients with negative pathology findings, 3% died as a direct consequence of prostate cancer, compared to 13% if the pathology was positive. Of the patients with positive seminal vesicle, 28% died of cancer. Conclusion Patients with any of the following factors have a risk of failure exceeding 40% and are candidates for studies of adjuvant therapy: seminal vesicle involvement, extension through the capsule, or margin involvement.