Overall five-year survival from breast cancer now approaches 85%. However, women with nodal involvement with or without metastatic disease have a much lower survival rate. Up to 20-30% of women may experience late relapse, presumably from latent cancer cells. Our goal is to design an assay that could detect circulating breast cancer cells. To accomplish our goal, blood is collected from breast cancer patients and normal volunteers. The samples go through a series of preparatory processes before performing quantitative real-time reverse transcriptase polymerase chain reaction (qRT-PCR) [Perkin-Elmer ABI Prism 7700]. In phase one of the preparatory process, the whole blood is run through a Ficoll-Hypaque gradient for the isolation of mononuclear cells. In the second preparatory process, half of the post-Ficoll specimen undergoes epithelial enrichment (EE) using immunomagnetic beads. Since breast cancer cells are epithelial cells, it is expected that amplification will be seen when fewer breast cancer cells are present in the blood that received EE than those without EE. RNA is extracted from the specimens with and without EE. The result of all the qRT-PCR runs will be analyzed together, and the specificity of our assay will be determined based upon the results from normal volunteers, while the cancer patient results will indicate the sensitivity of the assay. If the resultant assay has a sensitivity and specificity within the 85%-95% range, it may be useful to predict clinical outcomes for breast cancer patients.