Objective To investigate the feasibility of constructing a preeclampsia(PE)risk model based on multiple exosomal micrornas(miRNA)expression levels and to verify its efficacy in predicting PE.Methods A total of 1037 pregnant women who were archived in our hospital from June 2019 to December 2021 and whose gestational weeks were less than or equal to 20 weeks were selected as the research subjects.The expression of exosomal miRNA(including miR-155-5p,miR-215-5p,miR-203a-3p,miR-199a-5p and miR-125a-3p)in all samples was detected by qRT-PCR.Then,all patients were followed up to the end of pregnancy.The occurrence of PE during the follow-up period was counted,and all samples were divided into the PE group and the control group according to results.Cox regression was used to analyze the influencing factors of PE.The multi-miRNA risk model was constructed with ggrisk package,and the predictive effect of the model on PE was evaluated by receiver operating characteristic(ROC)curve.Results By the end of follow-up on October 31,2022,974 cases were finally followed up,and the follow-up completion rate was 93.92%.Among all the 974 patients who completed the follow-up,65 patients developed PE,so they were finally divided into the PE group,and 909 cases were used as the control group.The age,pre-pregnancy BMI and waist circumference at 12 weeks of gestation were higher in the PE group than those in the control group(P＜0.05).The proportions of smoking history and drinking history were higher in the PE group than those of the control group(P＜0.05).The contents of triglyceride(TG),low density lipoprotein cholesterol(LDL-C),total cholesterol(TC),alanyl aminotransferase(ALT),aspartate aminotransferase(AST),platelet distribution width(PDW),mean platelet volume(MPV),miR-155-5p,miR-199a-5p and miR-215-5p were higher in the PE group than those in the control group,while contents of thyroid stimulating hormone(TSH),miR-125a-3p and miR-203a-3p were lower in the PE group than those in the control group(P＜0.05).The expression levels of miR-125a-3p,miR-155-5p,miR-199a-5p and miR-215-5p were independent predictors of PE(P＜0.05).The predictive risk model constructed from the above miRNAs had good predictive value in the occurrence of PE(AUC=0.998),with a sensitivity of 98.46%(63/65)and a specificity of 93.94%(854/909).Conclusion miR-125a-3p,miR-155-5p,miR-199a-5p,miR-203a-3p and miR-215-5p are significantly related to the occurrence of PE,and the PE prediction model constructed with the above five miRNAs has better effect.