BACKGROUND: Gait disability affects the daily lives of patients with stroke in both home and community settings. An abnormal foot-ankle position can cause instability on the supporting surface and negatively affect gait. Our research team explored the ability of a portable peroneal nerve-targeting electrical stimulator to improve gait ability by adjusting the foot-ankle position during walking in patients with chronic stroke undergoing home-based rehabilitation. METHODS: This was a double-blinded, parallel-group randomized controlled trial. Thirty-one patients with chronic stroke and ankle-foot motor impairment were randomized to receive 3 weeks of gait training, which involved using the transcutaneous peroneal nerve stimulator while walking (tPNS group; n = 16, mean age: 52.25 years), or conventional home and/or community gait training therapy (CT group; n = 15, mean age: 54.8 years). Functional assessments were performed before and after the 3-week intervention. The outcome measures included spatiotemporal gait parameters, three-dimensional kinematic and kinetic data on the ankle-foot joint, and a clinical motor and balance function assessment based on the Fugl-Meyer Assessment of Lower Extremity (FMA-LE) and Berg Balance scales (BBS). Additionally, 16 age-matched healthy adults served as a baseline control of three-dimensional gait data for both trial groups. RESULTS: The FMA-LE and BBS scores improved in both the tPNS groups (p = 0.004 and 0.001, respectively) and CT groups (p = 0.034 and 0.028, respectively) from before to after training. Participants in the tPNS group exhibited significant differences in spatiotemporal gait parameters, including double feet support, stride length, and walking speed of affected side, and the unaffected foot off within a gait cycle after training (p = 0.043, 0.017, 0.001 and 0.010, respectively). Additionally, the tPNS group exhibited significant differences in kinematic parameters, such as the ankle angle at the transverse plane (p = 0.021) and foot progression angle at the frontal plane (p = 0.009) upon initial contact, and the peak ankle joint angle at the transverse plane (p = 0.023) and foot progression angle (FPA) at the frontal and transverse planes (p = 0.032 and 0.046, respectively) during gait cycles after 3 weeks of training. CONCLUSIONS: Use of a portable tPNS device during walking tasks appeared to improve spatiotemporal gait parameters and ankle and foot angles more effectively than conventional home rehabilitation in patients with chronic stroke. Although guidelines for home-based rehabilitation training services and an increasing variety of market devices are available, no evidence for improvement of motor function and balance was superior to conventional rehabilitation. Trial registration Chictr, ChiCTR2000040137. Registered 22 November 2020, https://www.chictr.org.cn/showproj.aspx?proj=64424.