To screen the key immune genes in the development of cervical cancer, construct immune related gene pairs (IRGPs), and evaluate their influence on the prognosis of cervical cancer. Tumor Genome Atlas (TCGA) database and geo database were downloaded as training set and validation set respectively, and immune related gene data were downloaded from immport. IRGPs model is established by machine learning, and the model is analyzed and evaluated. Using the Uclcan to analyze the immune genes expression in cervical cancer, and to further explore the association with the expression level and the clinical stage and prognosis of cervical cancer. According to the analysis of training set, we identified 29 IRGPs as key gene pairs and constructed the model. The AUC value of the model was greater than 0.9, and the model group survival rate was conspicuous different (P < 0.001). The reliability of the model was confirmed in the validation group. Our IRGPs play an important role in the occurrence and development of cervical cancer, and can be used as a prognostic marker and potential new target of cervical cancer.