Objective:This study aims to comprehensively evaluate and compare the effects of various exercise modalities on the control of chronic nonspecific low back pain through a network meta-analysis, in order to identify the most effective interventions. Methods: Adhering strictly to the PRISMA guidelines, this study selected randomized controlled trials from databases including PubMed, Web of Science, Embase, Cochrane Library, and Scopus, up to June 30, 2024. Data were processed using Stata 17.0 software, and the effect sizes were synthesized using standardized mean differences (SMD) with 95% confidence intervals (CI). The SUCRA method was used to rank the effectiveness of the interventions. Results: A total of 26 studies involving 1,507 participants aged between 20 and 63.5 years were included. The network meta-analysis revealed that yoga [SMD = -1.71 (-2.93, -0.49), P < 0.05] and core stability training [SMD = -0.81 (-1.44, -0.18), P < 0.05] were significantly more effective than the control group. SUCRA probability ranking indicated that Tai Chi (SUCRA = 77.4) might be the best modality for improving control of chronic nonspecific low back pain. Conclusion: This network meta-analysis demonstrates the intervention effects of different exercise modalities on chronic nonspecific low back pain, with Tai Chi potentially being the most effective intervention. This provides an important reference for non-pharmacological interventions in chronic nonspecific low back pain.