Objective To analyze the characteristics and significance of cell communication in the microenvironment of isocitrate dehydrogenase （IDH） mutation gliomas. Methods Seurat and clusterProfiler were used to analyze cell types， proportions， and functions in IDH mutation glioma tissues. CellphoneDB was used to analyze cell-cell communication characteristics and screen ligand-receptor （LR） pairs with significant differences. Univariate Cox regression was used to analyze the association between LR pairs expression and overall survival in IDH mutation glioma patients. Patients were classified into different subtypes using consistent clustering based on LR pairs expression and prognostic information. Log-rank， Gene Set Enrichment Analysis （GSEA）， ESTIMATE， and CIBERSORT were used to compare the survival differences， functional differences， and immune characteristics of patients with different subtypes. Based on the correlation of LR pairs expression and least absolute shrinkage and selection operator （LASSO） analysis， a risk prediction model was constructed to analyze the association between risk scores， patient subtypes and pathological indicators. Kaplan-Meier survival curve， Log-rank test and Cox regression model were used to evaluate the relationship between risk score and survival of IDH mutation glioma patients. Results A total of 27 154 cells were obtained， annotated and divided into six cell types. In the IDH mutation patients of CGGA693 and CGGA325 datasets， there were 140 and 131 LR pairs， respectively， that were associated with patient prognosis （P＜ 0.05）， and patients were divided into two molecular subtypes with significant differences in prognosis （P ＜ 0.05）， and the differences were statistically significant. Compared with subtype 1 with good prognosis， subtype 2 had a higher proportion of glioma patients with grade Ⅳ and 1p/19q co deletion （P＜ 0.05）. Hallmark enrichment analysis showed that subtype 2 was associated with MYC target genes， DNA repair， and oxidative phosphorylation. KEGG pathway enrichment analysis showed that subtype 2 was mainly associated with ribosomes， oxidative phosphorylation， Parkinson's disease， Huntington's disease， and Alzheimer's disease. Compared with subtype 1， subtype 2 had lower tumor purity and higher degree of immune infiltration. There were 14 LR pairs with expression correlation coefficients greater than 0.5 between ligands and receptors. A risk scoring model consisting of DLL4-NOTCH3， CXCL12-CR4， COPA-SORT1， PLAU-PLAUR， and EPHB2-EFNB1 was constructed using LASSO regression. LR pair risk score was higher in subtype 2 （P＜ 0.05）， which increased with pathological grade （P＜ 0.05）， and LR pair risk score was higher in the 1p/19q co deletion group （P＜ 0.05）， and the differences were statistically significant. In patients with IDH mutation， IDH mutation， and 1p/19q non deficient gliomas， the high-risk group had a shorter survival period compared to the low-risk group with a statistical difference （P ＜ 0.000 1）. Univariate and multivariate Cox regression analysis showed that WHO pathological grade （grade Ⅲ all P＜0.05； grade Ⅳ all P＜0.001）， 1p/19q status （all P＜0.05）， and risk score （all P＜0.001） were statistically significant factors affecting the overall survival of patients with IDH mutation gliomas. Conclusions LR pairs can serve as a predictive factor for molecular subtypes and prognosis in patients with IDH mutation gliomas