ObjectiveThe Hybrid Assistive Limb (HAL; CYBERDYNE, Inc., Japan) is a wearable robot device that provides effective gait assistance according to voluntary intention by detecting weak bioelectrical signals of neuromuscular activity on the surface of the skin. We used HAL for patients with amyotrophic lateral sclerosis (ALS) to determine whether HAL training had an effect on their gait ability. MethodsWe conducted a single-center, single-arm, observational study. Patients with ALS underwent HAL training once per day (20–40 min per session) for 9–10 days for at least 4 weeks. Gait ability was evaluated using the 2-minute walk test, the 10-meter walk test without the assistance of HAL, and activities of daily living (ADL) using the Barthel Index and Functional Independence Measures before and after a full course of HAL training. ResultsThere were no dropouts or adverse events during the observation period. Gait function improved after HAL training. The 2-minute walk test revealed a mean gait distance of 73.87 m (36.65) at baseline and 89.9m (36.70) after HAL training (p = 0.004). The 10-meter walk test showed significantly improved cadence, although gait speed, step length on the 10-m walk, or ADL measurements did not change significantly. ConclusionsAlthough HAL is not a curative treatment for ALS, our data suggest that HAL may be effective in ameliorating and preserving gait ability in patients with ALS.