Objective: To systematically evaluate the effectiveness and safety of vagus nerve stimulation (VNS) in treating treatment-resistant depression (TRD), and provide evidence-based support for clinical application. Methods: Following PRISMA guidelines, randomized controlled trials (RCTs) and pre-post controlled cohort studies on VNS for TRD published from January 2005 to June 2019 were retrieved from MEDLINE/PubMed, EMBASE, Web of Science, Cochrane Library, CNKI, and Wanfang databases. Literature quality was assessed using the Cochrane Risk of Bias Tool. Meta-analysis was performed with RevMan 5.0 software, with effect sizes represented by standardized mean difference (SMD) or risk ratio (RR). Heterogeneity was tested using the I 2 statistic, and sensitivity analysis was conducted to explore sources of heterogeneity. Results: A total of 7 studies (4 RCTs, 3 pre-post studies) involving 1,236 TRD patients (682 in the VNS group and 554 in the treatment-as-usual [TAU] group) were included. Meta-analysis showed that 12-month VNS treatment significantly reduced Montgomery-Åsberg Depression Rating Scale (MADRS) scores compared to baseline (SMD = -0.42, 95%CI = -0.61 to -0.23, P < 0.001), and the remission rate (MADRS ≤ 10) was significantly higher than in the TAU group (RR = 1.85, 95%CI = 1.32 to 2.59, P < 0.001). In terms of safety, the most common adverse events (AEs) with VNS were voice alteration (52.3%), cough (24.1%), and neck pain (18.7%), with most AEs resolving over time (e.g., dysphagia decreased from 13% to 4%). Conclusion: VNS demonstrates certain effectiveness in TRD, with long-term (≥12 months) treatment significantly improving depressive symptoms. It also exhibits good safety, with most AEs being tolerable.