The aim of this study was to compare Tc-99m sestamibi scintimammography and dynamic contrast-enhanced MR imaging for the evaluation of indeterminate mammographic lesions. Forty patients with questionable mammographic findings were included in a prospective study. Thirty lesions were non-palpable. Mean lesion size was 1.6+/-0.7 cm (range 0.5-3.5 cm). Scintigraphy was considered as malignant when focal tracer accumulation was present. In MR imaging, lesions were classified according to their signal intensity time course: no enhancement or steady enhancement with low signal intensity (M0); steady enhancement with high signal intensity (M1); or rapid enhancement with plateau (M2) or washout (M3). Lesions classified as M2 or M3 were considered as suspicious for malignancy. Histopathologic evaluation was performed in 24 lesions. In 16 cases lesions were classified as benign from follow-up examinations (mean 24 months). Malignancies were proven in 14 patients (9 invasive carcinomas, 5 ductal carcinoma in situ). Sensitivity of MR imaging was 12 of 14 (86%) and sensitivity of scintimammography was 8 of 14 (57%). One of 26 benign lesions was false positive at MR imaging. Scintigraphy showed no false-positive results. In conclusion, magnetic resonance imaging provided high accuracy in evaluation of indeterminate mammographic lesions. Sensitivity of scintimammography was too low in detecting small carcinomas.