The volleyball spike is repeated many times in practices and games, presenting a high risk of overuse injury. Previous biomechanical analyses estimating forces on the shoulder during spiking have not included the force exerted on the arm by the ball, because no practical method exists to estimate the contact force between the ball and the hand. The objective of the study was to model the internal shoulder joint reactions while including the measured ball contact force. Ten adolescent female volleyball players performed spikes while we recorded 3D motion capture data for both ball and player. Using an impulse-momentum analysis, we estimated the ball contact force, then included the force in a computational simulation model to estimate the torques produced by the shoulder. The study found that post-contact ball velocities range from 8.6 m/s - 18.2 m/s with net forces between 238 N - 672 N. Most notably, when the ball contact force was included, the average modeled internal shoulder torque to internally rotate the arm increased from -26 N-m to +44 N-m (p < 0.001). These data suggest that neglecting the contact force may risk misinterpreting connections between biomechanics and injury due to spiking. More accurate joint mechanics models will lead to better injury prevention recommendations for volleyball players of all ages.