Breast tumors are thought to originate, grow, and metastasize in an environment which includes steroid hormone receptors, their cognate steroid ligands, and many gene products which are regulated by steroid hormone receptor-ligand complexes. In this paper we describe highly sensitive and quantitative immunofluorometric procedures for measuring three proteins that are candidate prognostic indicators in breast cancer, namely, the p53 tumor suppressor gene product, carcinoembryonic antigen (CEA), and prostate specific antigen (PSA). These proteins were quantified in over 950 cytosolic tumor extracts along with estrogen and progesterone receptors (ER, PR). Association analysis between all five biochemical parameters revealed strong negative associations between p53 and receptors and strong positive associations between CEA and receptors. Negative associations between p53 and CEA and between CEA and PSA were also found. These associations, not quantitatively studied in previous reports, are related to each other using a hypothetical model. The observed associations may further contribute to the understanding of the biology of breast tumors.