Controversy exists about whether breaking pitches are more stressful than are fastballs. Previous biomechanical studies compared kinematics but not kinetics. Elbow and shoulder forces and torques are statistically different among the fastball, curveball, change-up, and slider. Descriptive laboratory study. Twenty-one healthy collegiate pitchers were studied with a high-speed automated digitizing system. All subjects threw fastballs (n = 21), most threw curveballs (n = 20) and change-ups (n = 19), and a few threw sliders (n = 6). Wrist, elbow, and shoulder kinetics were calculated using inverse dynamics. Nine kinetic and 26 kinematic parameters were compared among the different pitch types using repeated-measures analysis of variance. At the shoulder, internal rotation torque, horizontal adduction torque, abduction torque, and proximal force were significantly less in the change-up than in the other 3 pitches. Shoulder horizontal adduction torque was greater in the fastball than in the curveball and slider. Shoulder proximal force was greater in the slider than in the curveball. Elbow proximal force was less in the change-up than in the other 3 pitches. Elbow varus torque was greater in the fastball and curveball than in the change-up. Elbow flexion torque was greater in the curveball than in the change-up. The curveball and change-up demonstrated kinematic differences from the fastball, consistent with previous studies. There were significant kinematic differences between the fastball and curveball but few kinetic differences. The change-up had lower joint kinetics, lower angular velocities, and different body positions than the other 3 pitch types had. Results for the slider were inconclusive because of small sample size. Because the resultant joint loads were similar between the fastball and curveball, this study did not indicate that either pitch was more stressful or potentially dangerous for a collegiate pitcher. The low kinetics in the change-up implies that it is the safest.