The purpose of this study was to assess the preliminary effects and feasibility of a game-based rehabilitation program on improving upper limb (UL) function in people with stroke. Six individuals post-stroke received gamebased rehabilitation and intensive functional training to UL with the ArmAbleTM device, a novel interactive arm rehabilitative device, for at least 30 minutes/day, 5 days a week for 2 weeks. Baseline and post-intervention outcome measures included the Fugl-Meyer assessment-upper extremity (FM-UE) and action research arm test (ARAT). The feasibility outcomes included retention and adherence rate, safety, and the level of satisfaction following the intervention. With the Wilcoxon signed-rank test, there was a statistically significant difference in the ARAT score (0.036), but not for the FM-UE scores (0.058), post-intervention. We observed a 100% retention rate and 83.3% exercise adherence rate on completing the game-based rehabilitation program. All participants showed moderate to high level of satisfaction with the device, games, and the rehabilitation program. The game-based rehabilitation with the ArmAbleTM was feasible and effective on improving UL function post-stroke. All participants reported that training with the ArmAbleTM was fun, motivating and enjoyable.