Immunohistochemical analysis, the most efficient way of assaying for HER-2/neu overexpression in patients with invasive breast cancer, is subject to variation in sensitivity and evaluation when used in multiple laboratories. Cell lines with differing but constant levels of HER-2/neu expression have been advocated as standard material against which assay sensitivity can be gauged. Automated image analysis could provide more precise linear measurements of HER-2/neu expression than the subjective and categorical scoring system originally designed for the HER-2/neu clinical trials. Multiple European laboratories (range, 92-126) stained 7 cell line standards on 6 successive occasions using a variety of immunohistochemical assays for HER-2/neu. During the 2-year study period, a trend toward a standard sensitivity level was observed, with significant improvement in numbers of laboratories achieving appropriate results. Image analysis gave reproducible and significantly different linear values for a total of 621 HER-2/neu results on cell lines previously categorized manually as 3+, 2+, 1+, or 0. The use of cell line standards and image analysis have the potential to assist in standardizing immunohistochemical results for predictive markers and provide more accurate and quantifiable cut points for predicting clinical response to therapy, respectively.