: A diagnosis of breast cancer as early as possible is most important for a good prognosis. Among previous imaging modalities (mammography, sonography, DSA, thermography, CT) only mammography was able to prove a significant reduction of mortality, but unfortunately only for women above 50 years of age. Magnetic resonance mammography (MRM) is under technical and clinical evaluation for more than 8 years. Special surface coils and a variety of measurement sequences have been tested. By using "dynamic MRM", i.e., the repetitive imaging of the same slices before and at short time intervals after the injection of contrast medium, a high sensitivity (97.3%) and specificity (96.9%) was found in detecting breast cancers of different histology. In this study especially clinically or mammographically difficult cases were included in order to test the efficacy of dynamic MRM. Breast cancer can be proven or excluded with a very high degree of confidence. Conventional mammography had been yielding a vast number of false positive findings. With MRM, therefore, the number of biopsies can be substantially reduced. The remarkable uniformity of the maximum rate of signal enhancement in carcinomas may be explained by the changed vascularisation of malignant tumours based on early tumour angiogenesis. This change in vascularisation can be detected by combining the use of contrast medium with a high contrast sensitive measurement technique. This method seems to be much more reliable than mammographic or sonographic procedures. MRM needs further technical improvements. Since breast cancer is still the most frequent cause of cancer in women in the Western hemisphere, MRM should be used in all cases where is a discrepancy among radiographic, sonographic or clinical findings. However, MRM examinations should be performed only by experienced examiners, since a variety of pitfalls must be avoided.