Inertial release of the RCF-67 side release buckle is investigated in this thesis through review of the literature, finite element modeling, and experimental testing. A finite element (FE) model is developed using the explicit non-linear finite element code LS-DYNA. The propensity of the buckle to unlatch under various combinations of acceleration pulse amplitudes, durations, and belt tension is simulated and profiled. The non-linear modal response of the spring is found to be significant contributor. The results are correlated and compared against experimental test data within the literature. Rate sensitive foam is added to the buckle model to demonstrate prevention of inertial release through addition of a damping element. The FE model predicts the performance of the buckle with high fidelity and granularity both before and after the rate sensitive foam is added. Laboratory impact sled testing is performed on both configurations to further corroborate the results of the model. A "whipping" effect is observed in the buckle housing acceleration data and high speed video which causes oscillation in the acceleration of the housing, where peak accelerations are also higher and pulse durations are longer measured at the buckle CG compared to those measured on the rigid mounting structure. The wealth of experimental and computational data found in the literature, combined with that obtained in this study, allows for insightful conclusions to be drawn regarding the unlatching potential of the RCF-67 buckle as well as the effectiveness of adding rate sensitive foam to this buckle design. This modification to add damping provides a simple, inexpensive, and effective means of rendering a seat belt buckle safe under severe crash conditions where it may not otherwise perform as intended. Applicability to automotive seat belt buckle design in general is expanded on. LS-DYNA and the associated methodology used in this study proves effective in analyzing a buckle design for prevention of inertial release based on limited test data.