Background: Amplification/overexpression of c-erbB-2/Her-2 can predict the prognosis and the efficiency of certain chemotherapeutics, especially the Her-2-antagonist trastuzumab, in breast cancer. Whether immunohistochemistry(IHC) or fluorescence in situ hybridization(FISH) is more suitable for detecting amplification/overexpression is controversial. This study aimed to clarify this. Furthermore we wanted to decide if it is sufficient by FISH to only test for gene amplification, or if it is necessary to correct for the chromosome 17-number to see which tumours are amplified for c-erbB-2 and which are polysome. Methods: Cochrane and Pubmed databases were searched for reviews on IHC versus FISH in detecting overexpression/amplification of Her-2/c-erbB-2. Next; 26 biopsies from ductal carcinomas were marked for chromosome 17 with chromogenic in situ hybridization(CISH). Chromosome-number pr cell was calculated, compared to c-erbB-2-copies pr cell and to the tumours` IHC-positivity.Results: The literature on the subject was of poor quality. No reviews were systematically written. Most advocated testing by IHC primarily and testing IHC 2+-tumours with FISH. A number of others advocated FISH as the primary test. As to the specimens, five out of 26 specimens had non-concordant results on the number of c-erbB-2 copies compared to the Her-2/chromosome-17-ratio.Conclusion: A systematic review on the subject which compares FISH, CISH, IHC and their scoring systems is necessary. Furthermore the number of c-erbB-2-copies should be corrected by chromosome 17-copies when doing FISH. The lower limit of low-amplified tumours; when counting the number of c-erbB-2 copies; should perhaps be increased from 4 signals to 5 or 6 signals pr cell.