After stroke, gait training is a key component of rehabilitation, and most individuals use a variety of walking aids depending on their physical condition and environment. This study aimed to investigate the potential effect of a one-arm motorized gait device for gait assist of chronic hemiplegic stroke survivors through comparison with traditional gait devices (parallel bar and hemi-walker). This study was conducted on 14 chronic hemiplegic stroke survivors. The participants were asked to walk under three conditions using different gait devices, and their gait parameters during walking were collected and analyzed. The first condition involved walking on parallel bars; second condition, walking using hemi-walkers; and third condition, walking using one-arm motorized gait devices. With the use of a gait analysis system, the spatio-temporal gait parameters in each condition were collected, such as gait velocity, cadence, step length, stride length, single support time, and double support time. In the results by repeated-measures ANOVA or the Friedman test, a significant difference was found in the gait parameters among all three conditions (p< 0.05). The post-hoc test showed a significant change in the spatio-temporal gait parameters (especially, velocity, cadence and affected side single and double support time) when one-arm motorized gait device were used compared with parallel bars and hemi-walkers (p< 0.05). The results of this study suggest that one-arm motorized gait devices developed for hemiplegic stroke survivors may be more effective potentially than parallel bars and hemi-walkers in gait assistance of chronic hemiplegic stroke survivors.