Introduction: Chronic kidney disease is estimated to become the fifth leading cause of death globally by 2040. Due to the high incidence of fatigue in patients with end-stage renal disease without reliable pharmacological treatments, more and more studies on non-pharmacological interventions to improve physical function appear; which might be the best approach remains unknown. This study aimed to compare and rank the efficacy of all known non-pharmacological interventions on improving physical function from multiple outcomes for adults with end-stage renal disease. Methods: This systematic review and network meta-analysis included searches of PubMed, Embase, CINAHL, and Cochrane Library from inception to September 1, 2022, for randomized controlled trials of non-pharmacological interventions to improve physical function in adults with end-stage renal disease. Literature screening, data extraction, and quality appraisal were performed systematically by two independent reviewers. The frequentist random-effect network meta-analysis was adopted to pool the evidence from five outcomes, namely, 6-min walk test, handgrip strength, knee extension strength, physical component summary, and mental component summary, respectively. Results: A total of 1,921 citations were identified by this search, of which 44 eligible trials enrolled 2,250 participants, and 16 interventions were identified. All subsequent figures refer to comparisons with usual care. For increasing walking distance, the combined resistance and aerobic exercise with virtual reality or music were the most effective interventions, with a mean difference plus 95% confidence interval of 90.69 (8.92–172.46) and 92.59 (23.13–162.06), respectively. Resistance exercise with blood flow restriction (8.13, 0.09–16.17) was the best treatment to improve handgrip strength. Combined resistance and aerobic exercise (11.93, 3.63–20.29) and whole-body vibration (6.46, 1.71–11.20) were associated with improving knee extension strength. For life quality, all treatment effects did not show statistically significant differences. Conclusions: It was found via network meta-analysis that combined resistance and aerobic exercise is the most effective intervention. Besides, if virtual reality or music is added to the training, there will be better results. Resistance exercise with blood flow restriction and whole-body vibration might be good alternative treatments for improving muscle strength. None of the interventions improved quality of life, suggesting a need for alternative interventions in this regard. The results of this study contribute evidence-based data to decision-making.