Background: To provide reliable evidence to exercise rehabilitation therapists and clinicians, we compared and analyzed the effects of different exercise modalities on cardiopulmonary function in hemodialysis patients using Bayesian network meta-analysis. Methods: PubMed, OVID, Web of Science, Cochrane Library, Embase, Scopus, CINAHL, SPORT Discus, SinoMed, CNKI, Wanfang, and VIP were searched from inception to July 20, 2022. We included randomized controlled trials comparing 12 exercise modalities to improve cardiorespiratory fitness in hemodialysis patients. All statistical analysis was performed using STATA and R. Result: A total of 82 randomized controlled trials involving 4146 maintenance hemodialysis patients were included in this study. The pair-wise meta-analysis showed that all exercise modalities had a positive effect on all indicators of cardiorespiratory capacity. The network meta-analysis demonstrated that Blood flow restriction training (BFRT), Cycle exercise (CE), Inspiratory muscle training (IMT), Combined aerobic and resistance training (CT), and Aerobic training (AT) were significantly better than usual care for 6-min walkability; Medium intensity continuous training (MICT), CT, CE, and AT were considerably better than usual care for VO(2)Peak; body and mind training (MBT) and CT significantly improved SBP compared to usual care; and only MBT was significantly better than usual care for DBP. Both the two-dimensional plot and the radar plot demonstrated that CT had the best combined-effect on each index of cardiorespiratory fitness. Subgroup and sensitivity analyses demonstrated the robustness of the results. The evidence was mainly low to very low for this network meta-analysis. Conclusion: There is no one exercise that can achieve the best effect on all of the outcomes. The benefits of MBT in decreasing arterial blood pressure are unsurpassed by other exercise methods. The intervention effect of the CT is better and more stable. Electrical muscle stimulation training (MEST) can be employed in individuals who do not wish to exercise actively but may cause an increase in blood pressure. On the basis of the characteristics of different exercise types, guidelines developers, clinicians, and patients may employ them appropriately. Systematic review registration: https ://www.crd.york.ac.uk/PROSPERO/# record Details.