There is excess breast cancer mortality for African-Americans (AA) compared to European-Americans (EA) of 1.5-2.2 fold that first appeared in 1970s and has been worsening since. This disparity may not be explained solely by reduced access to medical care. We proposed that surgery to remove a primary tumor induces angiogenesis of distant dormant micrometastases in 20% of premenopausal node-positive patients. This hypothesis helps explain the reduced benefit of mammography for women aged 40-49. Interestingly, for AA the average age at diagnosis is 46 while for EA it is 57. The resultant increased proportion of AA premenopausal breast cancer suggests a possible explanation for the AA/EA excess mortality. Early detection, which began in the 1970s, is more effective in postmenopausal women than in premenopausal women. Since AA breast cancer is mostly premenopausal and EA breast cancer is mostly postmenopausal, it might be anticipated that starting in the 1970s because of surgery-induced early mortality, outcome would be superior for EA compared to AA.