Purpose: To evaluate whether relevant diagnostic information can be achieved when using magnetic resonance mammography (MRM) on midfield as a supplement to conventional imaging and clinical examination in women with primary breast cancer. Material and Methods: 30 women (55 breasts containing 49 malignant tumors) planned for uni or bilateral mastectomy were examined with dynamic MRM on midfield, 0.6T. The women were examined with mammography (M) and ultrasonography (US) prior to MRM. The descriptions of the conventional examinations were evaluated retrospectively, whereas the MRM was evaluated prospectively, with knowledge of the M+US findings. Imaging findings suggesting malignancy were registered and correlated with pathology after mastectomy. A home-made rating system for evaluation of the detected lesions was tested. Results: MRM detected seven additional malignant tumors, failed to detect three lesions and characterized four as gray-zone lesions according to the rating system. Sensitivity of finding the tumors with M+US was 79.0%, with a PPV for malignant tumors of 84.4%. One breast in which MRM found a malignant tumor had not initially been examined with US. Sensitivity with MRM was 91.6%, with a positive predictive value of malignant tumors of 97.7%. Conclusion: MRM on mid-field seems to improve the detection of cancers when used as a supplement to M+US in women with primary breast cancer. We believe that the results are fair compared to MRM on high-field, although further research and refinement are needed.