Objective:To analyze the risk factors of mycoplasma pneumoniae pneumonia(MPP)complicated with lobar pneumonia in children,and to construct its nomogram prediction model.Methods:The clinical data of 322 children with MPP were retrospectively analyzed.According to the imaging diagnostic criteria,they were divided into concurrent group(n=179)and non-concurrent group(n=143).The general data,laboratory indicators at admission,imaging features and treatment infor-mation of all children were collected.The differences of the above data between the two groups were compared.Multivariate Lo-gistic regression analysis was used to screen out the independent risk factors of MPP children complicated with lobar pneumoni-a,and the nomogram model was constructed by R software.The receiver operating characteristic(ROC)curve was drawn to verify the predictive efficacy of the model.Results:Of the 322 children,179 had lobar pneumonia,with an incidence of 55.59%(179/322).Compared with the concurrent group,the non-concurrent group was older(P<0.05),the proportion of fever dura-tion≥7 d,pleural effusion,pulmonary consolidation involving more than 2 lobes and the levels of PCT,D-dimer,CRP,ESR and LDH were higher(P<0.05),and the time from onset to application of macrolides was longer(P<0.05).Multivariate Lo-gistic regression analysis showed that the duration of fever≥7 days(OR=1.752,95%CI:1.324～2.319),high level of CRP(OR=2.433,95%CI:1.529～3.871),high level of LDH(OR=1.933,95%CI:1.377～2.713),pleural effusion(OR=3.504,95%CI:1.535～7.998)and long time from onset to application of macrolides(OR=2.591,95%CI:1.408～4.766)were independent risk factors for lobar pneumonia in children with MPP(P<0.05).The area under the curve(AUC)of the nomogram prediction model ROC was 0.920,the sensitivity was 85.47%,and the specificity was 87.41%.Conclusion:Persis-tent fever,CRP and LDH levels,pleural effusion and the use of macrolides from onset to onset are closely related to lobar pneu-monia in children with MPP.The nomogram prediction model constructed based on this has good predictive efficacy after inter-nal verification.