Objective：To analyze the clinical features and prognosis of non-small lung cancer（NSCLC）patients harboring SMARCA4 mutation，and detect the clinical significance of SMARCA4 gene via bioinformatics method. Methods：Clinical data from 35 patients with NSCLC harboring SMARCA4 mutations were retrospectively reviewed. According to the type of co-mutated gene，the patients were divided into several groups. The survival rate was calculated by Kaplan-Meier method. The expression of SMARCA4 in NSCLC tissue and normal lung tissue was investigated by TIMER database and GEPIA database. The cBioPortal database was used to analyze frequencies and types of SMARCA4 gene mutation，as well as the relationship of SMARCA4 mutation and prognosis. SMARCA4- related protein-protein interaction network was constructed by STRING database. Results：SMARCA4-mutated NSCLC patients were mainly old males，and lung adenocarcinoma was the main pathological type. The accompanying mutated genes included EGFR（14 cases），TP53（11 cases），KRAS（4 cases），ERBB2（4 cases），etc. Survival analysis found that the prognosis of patients with EGFR mutation was better than that of wild type，while the prognosis of patients with KRAS mutation was worse than that of wild type. Bioinformatics analysis showed that 8% of NSCLC patients carried SMARCA4 gene mutation，including missense mutations，splicing mutations and deletion mutations. In general，lung adenocarcinoma patients with SMARCA4 mutations had a poorer prognosis than wild- type patients. Conclusion：SMARCA4 mutations may affect the prognosis of NSCLC patients，and the accompanied mutations may have different effects on the prognosis，providing specific evidence for the precise treatment of lung cancer.