Background To investigate the effects of robot-assisted gait training (RAGT) on spasticity and pain in people with spinal cord injury (SCI). Material and methods. Four electronic databases (PubMed, Scopus, Medline, and Cochrane Central Register of Controlled Trials) were searched for studies published up to November 2019. Only human trials and of English language were included. The searched studies were reviewed and extracted independently by two authors. Randomized controlled trials (RCTs) and non-RCTs were pooled separately for analyses. Primary outcome measures included spasticity assessed by Ashworth scale (AS) or modified Ashworth scale (MAS) and pain assessed by VAS. Secondary outcome measures included lower extremity motor score (LEMS) and walking ability (i.e., 6-minute walk test, 10-meter walk test). Results A total of 225 studies were identified. Eighteen studies (7 RCTs and 11 non-RCTs) including 301 subjects met inclusion criteria. The outcome measure of spasticity significantly improved in favor of RAGT group in non-RCTs (AS: 95%CI = −0.202 to -0.068, p ≤ 0.001; MAS: 95%CI = −2.886 to -1.412, p ≤ 0.001). The results on pain did not show significant change after RAGT in either RCTs or non-RCTs. LEMS and walking ability significantly increased in favor of RAGT. Conclusions RAGT can improve spasticity and walking ability in people with SCI. The probable reason for no significant change in pain after RAGT is floor effect. RAGT is beneficial for normalizing muscle tone and for improving lower extremity function in people with SCI without causing extra pain.