Objective. This study aims to explore the therapeutic efficacy of San Zi Yang Qin Decoction (SZ) and its potential mechanism in the treatment of non-alcoholic fatty liver disease (NAFLD) based on network pharmacology and in vivo experiments. Methods. Effective chemicals and targets of SZ were searched in online databases, according to the drug-likeness of compounds and the binomial distribution of targets. A disease-target-chemical network was established using NAFLD-associated genes screened through GeneCards database, Gene Ontology (GO) terms, and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways. Furthermore, animal experiments were conducted to verify the efficacy and mechanism of SZ predicted by network pharmacology. The NAFLD mouse model was established with C57BL/6J mice fed with a high-fat diet for 22 weeks. The mice in the control group were fed with a chow diet. From the 23rd week, the NAFLD mice were treated with intragastric SZ or normal saline for 8 weeks. After the glucose tolerance was measured, the mice were sacrificed, followed by the collection of serum and liver tissues. Pathological changes in liver tissues were examined by H&E staining. Additionally, alanine aminotransferase (ALT), aspartate aminotransferase (AST), serum fast blood glucose, and insulin levels were detected. Expression levels of TNF-α of serum and liver tissues were determined by ELISA and qRT-PCR, respectively. Western blot was used to detect the activation of AKT in liver tissues. Results. A total of 27 effective compounds and 20 targets of SZ were screened. GO analysis uncovered a significant correlation between the targets of SZ and those of NAFLD. KEGG analysis presented the signaling pathways enriched in SZ and NAFLD, including NAFLD, TNF-α, and apoptosis pathways. The area under the curve of major GO and KEGG pathways indicated the potential role of SZ in improving NAFLD. In vivo experiments demonstrated that SZ significantly alleviated hepatosteatosis and inflammatory cell infiltration in liver tissues, reduced serum transaminases, and improved insulin resistance and glucose tolerance of NAFLD mice. The protein level of phospho-AKT was upregulated by SZ. Additionally, SZ treatment obviously impaired the TNF-α level in the serum and liver tissue of NAFLD mice. Conclusions. According to the network pharmacology analysis and in vivo experiments, SZ could have therapeutic efficacy for NALFD. The mechanism mainly involves pathways relative to insulin resistance, TNF-α, and apoptosis. Our results provide a scientific basis for SZ in the clinical treatment of NAFLD.