Systemic analysis of the relationship between the levels of methylation of 21 microRNA genes and the parameters of breast cancer progression was performed on a representative sample of 91 paired specimens of breast cancer and histologically normal tissues and a system of markers for prediction of metastasis was proposed. A significant association of hypermethylation of 11 genes with late (III-IV) clinical stages was found, and for 6 genes ( MIR124-1 , MIR127 , MIR34B/C , MIR9-3 , MIR1258 , and MIR339 ) this association was highly significant ( p ≤0.001, FDR=0.01). For MIR9-3 and MIR339 , an association with tumor size was demonstrated ( p <0.001, FDR=0.01). No association of the levels of methylation of the analyzed microRNA genes with the degree of differentiation were found. An association with lymph node metastasis was established for 9 microRNA genes; the most significant association was shown for 6 genes MIR125B-1 , MIR127 , MIR9-3 , MIR339 , MIR124-3 , and MIR1258 ( p <0.005, FDR=0.05). Based on these 6 genes, a marker system for predicting breast cancer metastasis was developed by ROC analysis. This system is characterized by 87% sensitivity and 77% specificity (AUC=0.894). The proposed system may have clinical application in the personalized treatment of breast cancer patients.