Introduction: The individual risk of a patient with coronary artery disease (CAD) for developing an acute coronary syndrome (ACS) is currently difficult to estimate. To date, there is no specific marker able to predict the risk for ACS. Objectives. We aimed to identify such a biomarker by studying the distribution of miRNA in 4 groups of patients: SA (stable angina, n = 22), UA (unstable angina, n = 21), MI1 (patients 1 month after myocardial infarction, n = 11), RF (subjects with CAD risk factors, but no significant CAD, n = 10), compared to C (healthy subjects, n = 10). Methods: MiRNA were isolated from pooled sera of SA, UA, MI1, RF and C individuals and following, a screening for 84 miRNA was performed. Fold change (FC) values of analyzed miRNA were expressed relative to C group. miR-125a, miR-146a, miR-33a (known to be associated with dyslipidemia, endothelial dysfunction and inflammation) were selected for validation in each patient’s serum using reverse transcription miRNA with stem loop specific primers and real time quantitative PCR with specific TaqMan primers. Results: Screening analysis showed that miR-122a (FC 63.07), miR-486a (FC 52.89), miR-92a (FC 51.19) had the highest values compared to C. All miRNA had higher serum levels in SA, UA, RF, MI1 groups, compared to C (p < 0.05 for all). MiR-486a and miR-92a had higher levels in patients with MI1 than UA. MiR-125a, miR-146a, and miR-33a did not differ between groups. Conclusions: In the analysed cohort the selected miRNA failed to discriminate between SA and UA. However, in more severe forms of CAD (ie MI1), levels of miR-486a and miR-92a were increased compared to UA.