Abstract Background and Aims Sleep disorders are commonly experienced by patients with end-stage kidney disease (ESKD) which influence the quality of life severely and may cause high risk of cardiovascular events. People with ESKD may constrain the use of neuroleptic medications which can lead to dependence if used in the longer-term. As non-pharmacological intervention, exercise training was suggested as based therapy to improve the quality of life. We conducted a systematic review to assess the efficacy of exercise training for improving sleep quality in dialysis patients with ESKD. Method Randomised controlled trials (RCTs) that evaluated the effects of regular exercise training on sleep quality in patients with ESKD were included. Searching was conducted in CENTRAL, EMBASE, PUBMED, CINAHL, WEB OF SCIENCE, and 3 Chinese databases (CBM, CNKI, WANFANG), date from inception to May 2019. Risk of bias was assessed. The quality and strength of evidence were evaluated by GRADE framework. Study selection, data extraction and methodological quality assessment were performed independently by two researchers. Results were reported as mean differences (MD) for continuous outcomes, with 95% confidence intervals (CI). This systematic review was registered in the PROSPERO (CRD42019141630). Results Twelve RCTs were included, which involving 753 participants. Methodological quality of studies and strength of evidence were assessed from low to moderate. Eleven trials included patients who receiving regular hemodialysis (HD) and one trial included patients that receiving either regular HD or continuous ambulatory peritoneal dialysis (CAPD). The major type of exercise was aerobic training during hemodialysis session (7 RCTs), followed by intradialytic resistance exercise, intradialytic mixed training, and interdialytic aerobic exercise. Most studies set exercise 3 times per week with moderate intensity, lasting 20∼60 minutes per session, and treatment duration for 2∼6 months. The Pittsburgh Sleep Quality Index (PSQI) questionnaire, Kidney Disease Quality of Life (KDQoL-SF) questionnaire, University of Massachusetts Weekly Sleep Questionnaire (UMass WSQ), visual analogue scale (VAS), and sleep fragmentation index (SFI) based on patient-reported outcomes were measured for testing sleep quality. By measuring with PSQI questionnaire (scale 0-21, lower is better), the sleep quality was improved with regular exercise comparing with usual care (5 RCTs, 296 participants, MD -4.92, 95%CI -6.54 to -3.30, I2=88%, moderate certainty evidence). The result was similar with KDQoL-SF questionnaire (4 items assessing sleep quality, higher is better) (3 RCTs, 328 participants, MD 8.75, 95%CI 2.06 to 15.43, I2=54%, low certainty evidence). However, there was no significant different in sleep diary by measuring with UMass WSQ (2 RCTs, 46 participants, MD -0.48, 95%CI -4.05 to 3.09, I2=59%, low certainty evidence). There was only one study measured with VAS and SFI, so we didn’t analyze it into meta-analysis. Conclusion Regular exercise appeared to be beneficial to improve sleep quality in patients with dialysis, especially intradialytic aerobic training. Given the growing challenge of pharmacological dependence, there is necessary to design high quality RCTs for non-pharmacological interventions such as exercise training.