Practice and accuracy of immunohistochemistry is known to vary highly. Reliability of HER-2 immunohistochemistry is critical because of its role in patient selection for therapeutical options in breast cancer. Therefore reliability of HER-2 immunohistochemistry in pathology laboratories in Austria was assessed. Ten tissue specimens of invasive ductal breast carcinomas and three cell line samples were tested. Presence/absence of gene amplification was determined by FISH to be used as a gold standard. Laboratories were asked to stain and assess slides using their routine immunohistochemical staining protocol. Overall the study consisted of 311 tests on tissue specimens and 142 on cell lines. In all cases manual scoring was performed. Participation was voluntary and was 94%. Overall sensitivity was 90.5% and specificity 99.2%. Overscoring including true false positive results were found in 6.7% and 6.3% in tissue specimens and cell lines, respectively. False negative determinations were obtained in 1.9% and 2.8% of tissue specimens and cell lines, respectively. HercepTest showed slightly higher reliability in comparison with individualized staining methods. By manual scoring inaccurate scoring affected 12.3% of test results and 62% of the laboratories. In conclusion participation rate and accuracy of HER-immunohistochemistry was high all over the country. Manually performed scoring demonstrated some limitations.