The authors evaluated 5623 cases of primary breast cancer followed for 1 to 21 years. Overall and breast cancer death rates were determined and compared to expected rates. Breast cancer patients showed overall and breast cancer death rates significantly higher than expected and which persisted at long-term follow-up. The observed/expected overall death ratios for follow-up periods of 0-5, 6-10, 11-15 or 16-20 years were 3.61, 2.55, 1.60 and 2.11, respectively. Death rates from breast cancer at 5, 10, 15 and 20 years were 20%, 32%, 40% and 48% respectively. The evidence of a persistent excess mortality even after long-term follow-up suggests the hypothesis that breast cancer is a systemic disease when clinically diagnosed. This study provided no evidence of a "clinical" cure for breast cancer patients. Even for N- patients the 5, 10, 15 and 20 year death rates from breast cancer were 12%, 20%, 28% and 38%, respectively. N- breast cancer, which is currently considered as a localized disease cured by surgery in most cases, would be better regarded to as a slow-growing metastatic disease, although "personal" cure may be achieved in many subjects dying of causes other than breast cancer.