Objective: To evaluate the state of evidence for the beneficial and harmful effects of Tai Chi on non-motor disorders in post-stroke patients. Design: Systematic review and meta-analysis of published studies. Subjects: Stroke survivors who received conventional rehabilitation therapy or Tai Chi training. Data sources: We searched seven electronic literature databases and one clinical registry platform to collect data from randomized controlled trials published up to July 26, 2020. Results: A total of 11 randomized controlled trials with 723 stroke survivors met the inclusion criteria, of which six were included in the meta-analysis. Among the 11 studies, one was assessed as "low", eight were assessed as "moderate", and only two were assessed as "high" for the assessment of methodologic quality. Compared to patients who received conventional rehabilitation therapy, those who received Tai Chi training showed greater improvement in scores of depression (standardized mean difference (SMD) [95% confidence interval (CI)] = 0.36 [0.10, 0.61], Grading of Recommendations Assessment, Development, and Evaluation [GRADE]: very low). There were no differences in the improvements in post-stroke global mental disorders (mean difference (MD [95% CI] = 6.15 [-3.05, 15.36], GRADE: moderate) or sleep disorders (MD [95% CI] = 0.33 [-1.51, 1.81], GRADE: low) between Tai Chi and control groups. Conclusion: Tai Chi may alleviate post-stroke depression in stroke survivors but has no clear effects on post-stroke cognitive and sleep disorders.