The aim of this study was to investigate whether the enumeration of circulating tumor cells (CTCs) in blood can differentiate between true localized and metastatic prostate cancer. A cross-sectional study of 104 prostate cancer patients with newly diagnosed high-risk prostate cancer was conducted. In total, 19 patients presented metastatic disease and 85 were diagnosed with localized disease. Analyses included intergroup comparison of CTC counts, determined using the CellSearch ® system, EPISPOT assay and GILUPI CellCollector ® , and ROC analysis verifying the accuracy of CTC count as a maker of disseminated prostate cancer. The vast majority (94.7%) of patients with advanced-stage cancer tested positively for CTCs in at least one of the assays. However, significantly higher CTC counts were determined with the CellSearch ® system compared to EPISPOT assay and GILUPI CellCollector ®. Identification of ≥4 CTCs with the CellSearch ® system was the most accurate predictor of metastatic disease (sensitivity 0.500; specificity 0.900; AUC (95% CI) 0.760 (0.613-0.908). Furthermore, we tried to create a model to enhance the specificity and sensitivity of metastatic prediction with CTC counts by incorporating patient's clinical data, including PSA serum levels, Gleason score and clinical stage. The composite biomarker panel achieved the following performance: sensitivity, 0.611; specificity, 0.971; AUC (95% CI), 0.901 (0.810-0.993). Thus, although the sensitivity of CTC detection needs to be further increased, our findings suggest that high CTC counts might contribute to the identification of high-risk prostate cancer patients with occult metastases at the time of diagnosis.