The accuracy in the diagnosis of metastatic colorectal cancer (mCRC) represents one of the challenges in the clinical management of patients. The detection of circulating tumour cells (CTC) is becoming a promising alternative to current detection techniques, as it focuses on one of the players of the metastatic disease and it should provide with more specific and sensitive detection rates. Here, we describe an improved method of detection of CTC from mCRC patients by combining immune-enrichment, optimal purification of RNA from very low cell numbers, and the selection of accurate PCR probes. As a result, we obtained a logistic model that combines GAPDH and VIL1 normalized to CD45 rendering powerful results in the detection of CTC from mCRC patients (AUROC value 0.8599). We further demonstrated the utility of this model at the clinical setting, as a reliable prognosis tool to determine progression-free survival in mCRC patients. Overall, we developed a strategy that ameliorates the specificity and sensitivity in the detection of CTC, resulting in a robust and promising logistic model for the clinical management of metastatic colorectal cancer patients.