Abstract: Methods of network pharmacology and bioinformatics were utilized to predict the mechanism of compound Muji granule of Manchu medicine in the treatment of liver cancer . The active ingredients and corresponding targets of compound Muji granule were collected by TCMSP and Swiss Target Prediction database. The related targets of liver cancer were screened through the GEO database. Venny 2.1 online software was used to obtain the common targets of drugs-disease, and then the “drug-compound-target-disease” network diagram was constructed by using the software Cytoscape 3.8.2. The STRING database was used to draw the protein-protein interaction network, and the perform GO function and KEGG pathway enrichment analysis were carried out through DAVID database. The relationship between the expression of key target genes and the survival curve was analyzed by using the Kaplan Meier-Plotter database, and the AutoDock platform was applied in verifying the molecular docking of active components and core protein targets. A total of 37 active components were obtained from compound Muji granule, and 57 common targets were screened, involving 98 biological processes, 17 cellular components, 37 molecular functions, and 15 signal pathways. The results of survival curve analysis show that the expression levels of ESR1, CYP3A4, and G6PD were related to the survival time of liver cancer patients. It was indicated that 6 active components and 6 key targets had a certain degree of binding through molecular docking. The active compounds in compound Muji granule, subprogenin C, naringenin, beta-sitosterol and wogonin, can act on targets such as ESR1, CYP3A4 and G6PD, which may play a therapeutic role in liver cancer by regulating p53, pentose phosphate, MAPK and other signaling pathways.