In the worldwide, uterine corpus endometrial carcinoma (UCEC) is the sixth most common malignancy in women, and the number of women diagnosed is increasing. Kinase plays an important role in the occurrence and development of malignant tumors. However, the research about kinase in endometrial cancer is still unclear. Here, we first downloaded the gene expression data of 552 UCEC patients and 23 healthy endometrial tissues from The Cancer Genome Atlas (TCGA), obtained 538 kinase-related genes from the previous literature, and calculated 67 differentially expressed kinases. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) were referenced to identify multiple important biological functions and signaling pathways related to 67 differentially expressed kinases. Using univariate Cox regression and Least absolute shrinkage and selection operator (LASSO), seven kinases (ALPK2, CAMKV, TTK, PTK6, MAST1, CIT, and FAM198B) were identified to establish a prognostic model of endometrial cancer. Then, patients were divided into high- and low-risk groups based on risk scores. Receiver operating characteristic (ROC) curves were plotted to evaluate that the model had a favorable predictive ability. Kaplan-Meier survival analysis suggested that high-risk groups experienced worse overall survival than low-risk groups. qRT-PCR and ISH assays confirmed the consistency between predicted candidate genes and real sample contents. CIBERSORT algorithm and ssGSEA were adopted to investigate the relationship between this signature and tumor immune microenvironment, and revealed that in low- and high-risk groups, the types of tumor-infiltrating immune cells and the immune cell-related functions were significantly different. In summary, a seven-gene signature risk model has been constructed, and could accurately predict the prognosis of UCEC, which may offer ideas and breakthrough points to the kinase-associated development of UCEC.