Our study aimed to confirm the therapeutic effects of using a trunk stabilization training robot (3DBT-33) in patients with chronic stroke. A total of 38 patients with chronic stroke were randomly assigned to either an experimental or a control group. The robot group (n = 19) received 30 min of trunk stability robot training in addition to conventional physical therapy, while the control group (n = 19) received a similar conventional physical therapy as the robot group. All participants were assessed using the following: the Functional Ambulation Categories (FAC), timed up and go (TUG) test, Berg Balance Scale (BBS), Korean Modified Barthel Index (K-MBI), and Fugl-Meyer Assessment of Lower Extremity (FMA-LE). There were statistically significant improvements in all parameters at follow-up assessment after 4 weeks of intervention (P < 0.05). There were statistically significant differences in the FMA-LE, K-MBI, and BBS between the robot and control groups (P < 0.05). There was no significant difference in FAC (P = 0.935) and TUG (P = 0.442). Minimal detectable change was more significantly observed in the FMA-LE and BBS than in FAC, TUG, and K-MBI. The findings in the present study showed that trunk stabilization rehabilitation training using a rehabilitation robot in patients with chronic stroke was effective in improving the balance and functions in the lower extremities.