p53, p185(erbB-2), and epidermal growth factor (EGF) receptor are well-characterized biomarkers in invasive adenocarcinoma of the breast and in ductal carcinoma in situ. erbB-2 Protein (p185(erbB-2)) must be identified clearly on cytoplasmic membranes while cytoplasmic expression is ignored for breast cancers to be classified as overexpressing p185(erbB-2). For p53, nuclear staining and the percent positive cells are considered, but rules for 'cut-offs' are not defined. Evaluation criteria for biomarkers in prostate cancer are preliminary and the 'rules' may not be the same as for breast cancer. Because the initial methods of fixation and tissue processing can change both the pattern and intensity of immunohistochemical identification of specific antigens and localization of receptors may change after the receptor-ligand interactions, we evaluated the effects of fixation both on the immunolocalization and intensity of expression of p53, p185(erbB-2), and EGF receptor. We also studied the patterns of p185(erbB-2) and p53 expression in a series of breast cancers evaluated concomitantly with a group of prostate cancers. Our results confirm that p53 mutations are common in breast cancer and that there is strong expression of p185(erbB-2) on the membranes of a subset of breast cancers. The patterns of staining for both p53 and p185(erbB-2) are different in prostate and breast cancers. A much lower proportion of prostate tumors than breast tumors have more than 10% of tumor cells with detectable nuclear p53 protein, but this proportion increases markedly in metastatic tumors or in primary stage D prostate cancer. In addition, a higher concentration of primary antibody is necessary to demonstrate nuclei showing p53 accumulation in prostate cancers than in breast cancers. The expression of p185(erbB-2) in prostate cancer is much coarser and more punctate than that observed in breast cancer; extensive cytoplasmic staining usually is detected in most prostatic adenocarcinomas. Concentrations of primary antibody to p185(erbB-2) that will stain membranes of breast cancers will not stain prostate cancer cells as strongly. However, the expression of p53 and/or p185(erbB-2) should not be ignored in prostate cancer since the rules for correlation of these biomarkers with the clinical outcome, or for use as surrogate endpoint biomarkers, have not been developed. Neutral buffered formalin is a poor initial fixative compared with other fixatives for demonstrating p53 accumulation or p185(erbB-2) expression. The effect of formalin fixation can be reversed with antigen retrieval methods, increasing detection markedly in cells with p53 accumulation.