Coronary artery disease (CAD) is the most prevalent cause of mortality and morbidity worldwide and the number of individuals at risk is increasing. To better manage cardiovascular diseases, improved tools for risk prediction including the identification of novel accurate biomarkers are needed. MicroRNA (miRNA) are essential post-transcriptional modulators of gene expression leading to mRNA suppression or translational repression. Specific expression profiles of circulating miRNA have emerged as potential noninvasive diagnostic biomarkers of diseases. The aim of this study was to identify the potential diagnostic value of circulating miRNA with CAD. Circulating miR-145, miR-155, miR-92a and let-7c were selected and validated by quantitative PCR in 69 patients with CAD and 30 control subjects from the cross-sectional study GENES. The expression of miR-145, miR-155 and let-7c showed significantly reduced expression in patients with CAD compared to controls. Multivariate logistic regression analysis revealed that low levels of circulating let-7c, miR-145 and miR-155 were associated with CAD. Receiver operating curves analysis showed that let-7c, miR-145 or miR-155 were powerful markers for detecting CAD. Furthermore, we demonstrated that the combination of the three circulating miRNA managed to deliver a specific signature for diagnosing CAD.