CANCER cells can enter the blood stream. Some of these circulating tumor cells (CTC) extravasate and form distant metastases which ultimately lead to the death of the patient. CTC have been observed quite some time ago (1-3). The major difficulty for the identification of CTC is their extremely low frequency (~1/ml in metastatic patients). When no CTC were detected in a tube of blood of a patient with metastases one always wonders whether they were not present or whether they were missed by the detection technology. Another hurdle is the extensive heterogeneity of the physical and immunological appearance of CTC within and between patients.Tumor cell lines play a pivotal role in both informing technology developers of the probable physical and immunological properties of CTC, as well as in the determination of the analytical accuracy, reproducibility, and linearity of any developed technology. However, tumor cell lines are not CTC. Excellent performance on tumor cell lines does not warrant that (1) the developed technology will actually identify CTC in blood samples of cancer patients nor that (2) these CTC relate to clinical outcome nor that (3) information relevant for treatment can be extracted.At the start of the trajectory toward the FDA cleared CellSearch system very little was known about the CTC frequency and their physical and immunological appearance. CTC enrichment in 10 ml of blood was performed by manual immunomagnetic enrichment targeting EpCAM using the GA73.3 antibody. The enriched cell suspension was fluorescently labeled with the PE-labeled CAM5.2 antibody recognizing cytokeratins 7 and 8, a PerCP-labeled antibody recognizing CD45 and a nucleic acid dye that could be measured in the FITC channel of a flowcytometer. Using this approach, it was first shown that CTC indeed could be detected in 10 ml blood of cancer patients with metastatic disease and at a lesser frequency in patients with no detectable metastasis while few "CTC" were detected in healthy controls (4-8). These results were sufficiently encouraging to further develop the CTC technology and design and conduct clinical studies to explore the clinical utility of CTC. The commercial CellSearch system that was ultimately developed included (1) a blood collection tube to preserve the fragile CTC for a period of up to 96 h, (2) an automated sample preparation system in which (3) the EpCAM antibody was replaced by VU1D9, (4) CAM5.2 cytokeratin antibody was replaced by C11 and A.53B/A2 to increase the range of cytokeratins, (5) the CD45-PerCP was switched to CD45-APC, (6) the nucleic acid dye was switched to DAPI, and (7) the flow cytometer replaced by a semi-automated fluorescence microscope with dedicated software to present CTC candidates to a reviewer. Thanks to an excellent team of reagent developers and engineers the system was reliable and provided accurate results over a large range of spiked tumor cells in blood with a high sensitivity and specificity. Whereas few "CTC" were detected in healthy donors and patients with benign disease...