Mammography is the screening test of choice for breast cancer. Its low specificity leads to a large number of unnecessary biopsies. Scintimammography, with either Tc-sestamibi (MIBI) or Tc-anti-carcinoembryonic antigen (CEA) Fab', has been proposed as a non-invasive test to lower the high false positive rate of mammography in certain patients. The two agents have not been compared, nor has their combined application been evaluated. We performed a prospective, non-randomized, open-label, single-centre study of 32 women with clinically and mammographically suspected breast cancer [Breast Imaging Reporting and Data System (BI-RADS, American College of Radiology) 4 or 5]. All patients underwent Tc-MIBI and Tc-anti-CEA Fab' scintimammography, and the results were correlated with histopathology. Overall, the accuracies for MIBI and CEA scans were 90.3% (28/31) and 77.4% (24/31), respectively. The probability of disease after mammography was 0.939+/-0.081 (95% confidence interval, CI). The post-mammography probabilities after positive MIBI or CEA scan were 0.965 and 0.960, respectively, and after negative MIBI or CEA scan 0.750 and 0.875, respectively. None of the above differences is significant. The post-test probability when both scans were positive (irrespective of which was performed first) was 0.977. It can be concluded that there are indications that scintimammography with Tc-MIBI is superior to that with Tc-anti-CEA Fab' when these tests are used as screening tests for breast cancer. However, mammography remains the screening test of choice for highly suspicious clinically palpable breast lesions. In this group of patients, the application of scintimammography with either Tc-MIBI or Tc-anti-CEA Fab' (alone or in combination) offers no additional advantage.