Hand strength weakness affects the performance of most activities of daily living. This study aims to design, develop, and test an electromyography (EMG) biofeedback training system based on serious games to promote motivation and synchronization and proper work intensity in grip exercises for improving hand strength. An EMG surface sensor, soft balls with different stiffness and three exergames, conforms the system to drive videogame clues in response to EMG-inferred grip strength, while overseeing motivation. An experiment was designed to study the effect of performing handgrip (HG) exercises with the proposed system versus traditional exercises. Participants, organized into two groups, followed a training program for each hand. One group followed a HG exergame training (ET) with the dominant hand and traditional HG training with the nondominant hand and inverse sequence by the second group. Initial and final grip forces were measured using a digital dynamometer. Questionnaires evaluated motivation and user experience, and exercise performance was evaluated in terms of work and rest time percentage and maximal voluntary contraction percentage over contraction periods. Data were analyzed for statistically significant differences and increase of means. Participants showed significantly better exercise performance and higher grip forces, with sustained intrinsic motivation and user experience, with the ET. Improvement in force level arises evidently from the synchronized work-rest time pattern and appropriated intensity of the muscle activity. This leads to support that EMG biofeedback exergames improve motor neurons firing and resting.