Objective:To investigate the relationship between serum vascular endothelial growth factor (VEGF), peripheral blood microRNA-126 (miR-126) and the number and distribution of cerebral microbleeds (CMBs).Methods:Consecutive patients with non-acute ischemic cerebrovascular disease admitted to the Department of Neurology, the First Affiliated Hospital of Baotou Medical College from June 2019 to June 2020 were enrolled. The clinical data were collected, 3.0 T MRI examination was performed, and susceptibility-weighted imaging was used to detect CMBs. The serum VEGF concentration was detected by enzyme-linked immunosorbent assay, and miR-126 was detected by fluorescence quantitative polymerase chain reaction. Multivariate logistic regression analysis was used to determine the independent influencing factors of CMBs. Multiple linear regression analysis was used to determine the correlation between serum VEGF concentration, miR-126 in peripheral blood and the number of CBMs. Receiver operating characteristic (ROC) curve was used to evaluate the predictive value of serum VEGF concentration and relative expression of miR-126 in peripheral blood for CMBs. Results:A total of 193 patients with non-acute ischemic cerebrovascular disease were enrolled, including 110 patients (57.0%) in the non-CMBs group, 20 (10.4%) in the strictly lobar CMBs group and 63 patients (32.6%) in non-strictly lobar CMBs group. The comparison among the three groups showed that age might be a risk factor for strictly lobar CMBs, while higher VEGF, higher cystatin C level, lower relative expression of miR-126 in peripheral blood, hypertension and previous stroke or transient ischemic attack might be the risk factors for non-strictly lobar CMBs. Multivariate logistic regression analysis showed that higher serum VEGF concentration was an independent risk factor for non-strictly lobar CMBs (odds ratio 1.186, 95% confidence interval 1.035-1.358; P=0.014), while the higher relative expression of miR-126 was an independent protective factor for non-strictly lobar CMBs (odds ratio 0.154, 95% confidence interval 0-0.269; P=0.026). Multiple linear regression analysis showed that higher serum VEGF concentration ( r=0.848, P<0.001) and the lower relative expression of miR-126 ( r=-0.043, P=0.035) significantly increased the number of CMBs. ROC curve analysis showed that the area under the curve of serum VEGF for predicting non-strictly lobar CMBs was 0.803 (95% confidence interval 0.741-0.865), the optimal cut-off value was 120.55 ng/L, the sensitivity was 70.7%, and the specificity was 75.5%. Conclusions:In patients with non-acute ischemic cerebrovascular disease, there is a significant correlation between serum VEGF concentration and the relative expression of miR-126 in peripheral blood and the number and distribution of CMBs. Serum VEGF can be used as a biomarker for predicting the presence of non-strictly lobar CMBs.