Scintimammography using Tc-99m sestamibi and contrast enhanced MRI were performed in order to determine the accuracy of both methods in the diagnostic work up of patients with suspicious or indeterminate preliminary diagnosis. 25 controls and 56 patients (14 with suspicious and 42 with indeterminate preliminary diagnoses), in whom physical examination and/or mammography warranted breast biopsy, underwent prone planar scintimammography. Sestamibi uptake was scored visually and measured using the ROI technique to enable semiquantitative evaluation. The patient group additionally underwent plain and contrast enhanced MRI. Visually determined signal increase following application of Gd-DTPA was compared with scintigraphic findings and final histopathologic results. Sensitivity and specificity of semiquantitative scintimammography for diagnosing breast cancer was 88% and 87%, respectively. Based on ROC analysis a target/non target ratio R > 1.3 was shown to be the optimal threshold for separating benign from suspicious scintigraphic diagnoses. MRI reading provided a slightly higher sensitivity (91%), but a considerable lower specificity (52%) due to contrast enhancement of different benign lesions. In the clinically important patient subpopulation with indeterminate results from previous diagnostic procedures, sensitivity of scintimammography fell to 79%, while specificity remained at 87%. MRI revealed a higher sensitivity of 89% and a lower specificity of 52%. Our data indicate that semiquantitative scintimammography using Tc-99m sestamibi provides a comparable sensitivity to contrast enhanced MRI in the assessment of breast cancer. The latter does not reduce the number of biopsies yielding benign results due to the high number of false positive diagnoses. Therefore, scintimammography seems to be the preferable tool in the diagnostic work-up of patients with indeterminate mammographic diagnoses.