Purpose The aim of this study was to investigate the effects of robotassisted gait training on proprioception, balance, and walking ability in stroke patients. This study also aimed to apply more effective treatments in clinical settings.  Methods A total of 24 stroke patients were divided into two groups: an experimental group receiving robotassisted gait training, and a control group receiving treadmill training. Both interventions were conducted for 30 minutes per session, three times a week, over four weeks.  Results In the proprioception assessment, the experimental group's joint position sense score decreased from 10.61±6.06 to 5.86±1.85, while the control group's score decreased from 5.60±3.09 to 4.28±2.85. No significant differences were found within or between the groups (p>.05). In the balance assessment, there were no significant changes in either group after the intervention (p>.05). Walking ability improved in both groups, but the control group showed a significant increase from 32.17±14.90 to 40.92±14.20 (p<.05), with a significant difference between the two groups (p<.05).  Conclusion Both robotassisted gait training and treadmill training did not significantly improve proprioception or balance in stroke patients. However, the control group showed significant improvements in walking ability. This suggests that treadmill training may be more effective for improving walking ability, while robotassisted training may have limited impact on proprioception and balance. Future research should focus on longer interventions and more precise assessments of proprioception's role in recovery.