Normal walking includes coordinated and controlled movement of the legs and arms. However, patients following stroke often present with inappropriate motor control which limits coordinated movement patterns of the affected limbs. This study aimed to compare the effects of rhythmic arm swing and arm fixation during treadmill walking in patients with poststroke hemiparesis. We used an alternating study design with multiple baselines across subjects. Three patients with chronic stroke participated in this study. During treadmill walking, rhythmic arm swing and arm fixation conditions were alternately applied. Outcome measures included the 10-meter walk test (10MWT) and energy expenditure index (EEI). In the intervention phase, all subjects showed significantly greater improvements in the 10MWT and EEI scores for rhythmic arm swing condition compared to those for arm fixation condition (p < 0.05). 10MWT improvement rates: Subject 1-34.81% vs. 15.75%; Subject 2-40.00% vs. 17.95%; and Subject 3-38.08% vs. 21.85%; and EEI improvements: Subject 1-23.19% vs. 14.08%; Subject 2-26.15% vs. 20.43%; and Subject 3-22.99% vs. 14.49%. These findings suggest that rhythmic arm swing is clinically feasible as a more favorable option to enhance the effects of treadmill walking training. However, larger studies with a different study design are needed to be able to make any judgment about the usefulness of the treatment.