Abstract: The major objective of this dissertation is to develop a model based on understanding of the injury mechanisms, injury trends, and injury potential to motorcyclists in certain types of motorcycle collisions, using computational models of the collision process. MADYMO (MAthematical DYnamic Model), a commercially available software program, is used to construct and analyze a realistic biodynamic model in an effort to understand the dynamics of the motorcycle-rider system. Modification strategies that may effectively reduce trauma to motorcyclists are also investigated. These modifications mainly involve identifying motorcycle geometry that may be hazardous to the rider during a crash. A study of this nature is very unique and provides new data to the scientific community. The MADYMO simulations are performed for the purpose of determining motorcycle rider injury potential based on the geometry of the motorcycle design. Validation of the models was previously performed (motorcycle validation by Neiboer et al. and MATD validation by ISO 13232) using proven scientific biomechanical and collision analysis procedures. Due to the limitations of the MADYMO motorcycle and MATD models, modifications were limited to changes in Delta-V, handlebar location, windshield presence, and gas tank orientation. Individual simulations were run with and without the geometric modification in place. Corresponding injury criteria values were calculated. Statistical analysis was then performed to determine if the change in motorcycle geometry, at different impact severities, was significant. Delta-V correlated significantly (p < .05) to all the injury parameters measured (Gadd Severity Index-GSI, Head Injury Criterion-HIC, Femur Force Criterion-FFC, Tibia Index-TI, Contiguous 3 ms, and Pelvic Force). The most interesting finding regarding gas tank orientation and pelvic force came from the exponential regression analysis. The average pelvic force increased exponentially (R2 >.99) as the gas tank angle increased. Statistically, windshield presence (or absence) correlated significantly with GSI and HIC. Although, graphically, the windshield appeared to have an affect on TI, it was not statistically significant. Handlebar location was not statistically significant with the injury parameters measured.