Seven right-handed teaching professionals and eight intermediate tennis players were filmed using two high-speed cameras (100 Hz) as they performed open and square stance forehand drives. Three-dimensional coordinates (3D) were reconstructed using the DLT method. A three-segment rigid body model of the racket and upper extremity was used to calculate the kinetics of the wrist, elbow, and shoulder joints up to impact. The open stance created lower resultant velocities of the racket at impact (21.2 and 15.8 m/s) than the square stance (22.3 and 16.4 m/s) for professional and intermediate subjects, respectively. The largest components of the resultant joint torques were generated by the shoulder horizontal adductors, followed by elbow varus torques, and shoulder internal rotation torques. Torques were similar across stance and skill level except for significantly (p < 0.05) greater peak shoulder internal rotation torques in the square compared to the open stance, greater peak wrist flexion torques in the intermediate compared to the professionals, and greater peak wrist flexion torques in the square stance compared to the open stance. The data did not support the hypothesis that the open stance technique creates greater loading throughout the upper extremity than the square stance technique. Peak upper extremity torques were similar to peak torques reported for baseball pitching and represent loads that could contribute to strength imbalances and overuse injuries.