Measuring HER-2 is important for selecting optimal therapy and predicting prognosis in breast cancer patients. Current methods for evaluating HER-2 include measuring protein overexpression by immunohistochemistry (IHC), measuring gene copy number by fluorescent in situ hybridization (FISH), and measuring shed antigen in the serum by enzyme-linked immunosorbent assay (ELISA). This review compares these 3 methods and analyzes the current literature pertaining to this subject. In comparing IHC with FISH, the negative predictive value is excellent for commonly used commercial antibodies but the positive predictive value is highly variable. However, by considering only strongly staining cases as positive by IHC, the positive predictive value is markedly improved. ELISA is useful in the follow-up care of patients with breast cancer. An algorithm for using all 3 methods is presented.