Purpose: PSA testing results in unnecessary biopsy and over-diagnosis with consequent overtreatment. Tissue biopsy is an invasive procedure, associated with significant morbidity. More accurate non-or minimum-invasive diagnostic approaches should be developed to avoid unnecessary prostate biopsy and over-diagnosis. We investigated the potential of using circulating tumor cell analysis in cancer diagnosis, particularly in predicting clinically significant prostate cancer in pre-biopsy patients. Material and methods: We enrolled 155 treatment naïve prostate cancer patients and 98 pre-biopsy patients for circulating tumor cell numeration. RNA was extracted from circulating tumor cells from 184 patients for gene expression analysis. Kruskal-Wallis, Spearman's rank, multivariate logistic regression and random forest were applied to assess the association of circulating tumor cells with aggressive prostate cancer. Results: In localized prostate cancer patients, 54% were scored as circulating tumor cell positive, which was associated with higher Gleason score (p=0.0003), risk group (p<0.0001) and clinically significant prostate cancer (p<0.0001). In pre-biopsy group, positive circulating tumor cell score in combination with PSA predicted clinically significant prostate cancer with AUC=0.869. A 12-gene panel prognostic for clinically significant prostate cancer was also identified. Combining PSA level, circulating tumor cell-score and the 12-gene panel, AUC for clinically significant prostate cancer prediction was 0.927 and in cases with multi-parametric MRI data, adding these to multi-parametric MRI significantly increased the prediction accuracy (AUC 0.936 vs 0.629). Conclusions: Circulating tumor cell analysis has the potential to significantly improve patient stratification by PSA and/or multi-parametric MRI for biopsy and treatment.