The purpose of this study was to better understand how lower body kinematics relate to peak glenohumeral compressive force and develop a regression model accounting for variability in peak glenohumeral compressive force. Data were collected for 34 pitchers. Average peak glenohumeral compressive force was 1.72% ± 33% body weight (1334.9 N ± 257.5). Correlation coefficients revealed 5 kinematic variables correlated to peak glenohumeral compressive force (P < .01, α = .025). Regression models indicated 78.5% of the variance in peak glenohumeral compressive force (R2 = .785, P < .01) was explained by stride length, lateral pelvis flexion at maximum external rotation, and axial pelvis rotation velocity at release. These results indicate peak glenohumeral compressive force increases with a combination of decreased stride length, increased pelvic tilt at maximum external rotation toward the throwing arm side, and increased pelvis axial rotation velocity at release. Thus, it may be possible to decrease peak glenohumeral compressive force by optimizing the movements of the lower body while pitching. Focus should be on both training and conditioning the lower extremity in an effort to increase stride length, increase pelvis tilt toward the glove hand side at maximum external rotation, and decrease pelvis axial rotation at release.