To provide new ideas for further research and treatment of nonalcoholic fatty liver fibrosis, we used bioinformatics technology to search the gene microarray data related to this disease and identified differentially expressed genes and potential therapeutic drugs. Gene Expression Omnibus (GEO) was used to search the entry of "nonalcoholic fatty liver fibrosis"; the GSE109345 microarray data was downloaded, the differentially expressed genes in the control group and the fibrosis model group were screened with the BioJupie analysis platform, and GO function, KEGG pathway, protein-protein interaction (PPI) network analysis and visualization were conducted for the differentially expressed genes. Finally, through the Connectivity Map (CMap) data platform, compounds with potential efficacy on nonalcoholic fatty liver fibrosis were predicted. A total of 109 differentially expressed genes were screened, including 70 up-regulated genes and 39 down-regulated genes. Functional analysis showed that differentially expressed genes were mainly involved in protein kinase B signal transduction, extracellular domain function, small molecule binding and other functions; pathway analysis showed that these genes participated in retinol metabolism, steroid hormone synthesis, and arachidonic acid metabolism; PPI network analysis showed that metallopeptidase inhibitor 1 (TIMP1), chemokine (C-C motif) ligand 2 (CCL2), recombinant signal regulatory protein beta 1 (SIRPB1), and cytochrome P450 (CYP) were the main genes related to nonalcoholic fatty liver fibrosis. CMap analysis showed that pioglitazone, midodrine, arecoline and other compounds had potential efficacy in nonalcoholic fatty liver fibrosis. Thus, by screening for differentially expressed genes, related genes and potential therapeutic compounds effective in the treatment of non-alcoholic fatty liver fibrosis can be identified, as well as new ideas and approaches for the clinical treatment of non-alcoholic fatty liver fibrosis.