Detection of residual tumor cells in BM and PBPC products has been correlated with worse outcome of breast cancer patients. Still, there is a considerable demand for studies investigating the influence of the actual tumor cell number on prognosis, as quantification routinely has been cumbersome and time consuming and, thus, was evaded. We developed and evaluated a competitive RT-PCR-ELISA assay for cytokeratin 19 (CK19) with standard curve quantification that allows quantification of multiple samples within a working day; mRNA isolation, RT-PCR reaction, and automated ELISA detection were carried out using commercial kits. Results were expressed as OD420nm ratios of CK19 and an internal competitor. Values were then converted into tumor cell numbers using a standard curve of MCF-7 tumor cells. The assay had high specificity because of primers and capture probes with great heterogeneity to both published pseudogenes, which was confirmed by BLAST sequence alignment. We achieved a sensitivity of detecting 1 tumor cell per 10(6) mononuclear cells (MNC). Between-batch precision (n = 8) for quantification was consistent and reasonable, with a coefficient of variation around 25%. Therefore, this assay should be suitable and sufficient for routine quantification of tumor cell numbers in BM or PBPC samples.