Objective To verify the core functional component group (CFCG) and potential pathways of Erzhi Pill (EZP) in treating liver fibrosis based on network pharmacology and experimental analysis.Methods All chemical components of EZP were collected from TCMSP database and literature. The PreADMET, Similarity Ensemble Approach and SwissTargetPrediction were used to filter potential active components and predict targets. GeneCards, DisGeNET and STRING were used to collect pathogenic genes and construct protein-protein interactions (PPI) networks, and weighted gene networks were constructed by Cytoscape 3.9.1. A new model was built to construct the functional effect space (FES), and cumulative contribution rate (CCR) of components was then calculated to obtain CFCG, and finally GO and KEGG pathway enrichment analyses were performed. The in vitro cultured human hepatic stellate cells (LX-2) cells were divided into the blank group (no TGF-β1 stimulation), the control group (20 ng·mL^-1 TGF-β1 stimulation) and the compound groups (0, 6.25, 12.5, 25, 50, 100, 200 μmol·L^-1). CCK-8 assay was performed to detect the effects of the compounds on the viability of the cells, and qPCR assay was used to detect the effects of compounds on the expression of collagen type IA1 (COL1A1) in LX-2, and Western blotting assay to verify the potential therapeutic mechanism.Results The CFCG of EZP contains a total of 22 compounds. According to their CCR ranking, the top eight compounds were selected for CCK-8 assay, and all of them were not cytotoxic within the tested concentration. The qPCR results showed that protocatechuic acid, quercetin, cynaroside, and luteolin significantly inhibited COL1A1 expression compared with TGF-β1-stimulated cells. GO and KEGG analysis and Western blotting validation revealed that these four compounds inhibited the expressions of PI3K, p-PI3K, AKT, p-AKT, and HIF1-α.Conclusion Protocatechuic acid, caffeic acid, quercetin, cynaroside and luteolin are the core components of EZP to exert anti-hepatic fibrosis effects, and its mechanism may be related to the regulation on PI3K-AKT and HIF1-α signaling pathways.