Current National Aeronautics and Space Administration (NASA) occupant protection standards and requirements are based on extrapolations of biodynamic models, which were based on human tests performed during pre-Space Shuttle human flight programs. In these tests, occupants were in different suit and seat configurations than are expected for the Multi-Purpose Crew Vehicle (MPCV) and Commercial Crew programs. As a result, there is limited statistical validity to the occupant protection standards. Furthermore, the current standards and requirements have not been validated in relevant spaceflight suit and seat configurations or expected loading conditions. The objectives of this study are to use current modeling techniques and industry standard approaches to analyze existing human databases to develop new NASA standards and requirements for occupant protection. To accomplish these objectives we began by determining which critical injuries NASA needs to mitigate. We then defined the anthropomorphic test device (ATD) and the associated injury metrics of interest. Finally, we conducted a literature review of available data for the Test Device for Human Occupant Restraint (THOR) ATD to determine injury assessment reference values (IARV) to serve as a baseline for further development. The first objective of this study was to assess the THOR for use by NASA in developing and validating occupant protection standards. THOR was impact tested in various orientations and multiple peak acceleration levels, and responses to these impacts differed from human responses under identical impact conditions and do not replicate human bracing for impact. As part of this study, a finite element model (FEM) of the THOR was developed, calibrated, and optimized. The FEM proved a good match to the ATD. The FEM simulations (at a wide range of impacts that military subjects participated in) historically revealed that the FEM correlated to lower-peak acceleration impacts better than to higher-peak accelerations. The second objective of this study was to mine existing human injury and exposure data. The U.S. Air Force maintains thousands of human-impact-test results. Data collected from 1976 to 2004 have been downloaded and converted into a user-friendly MATLAB data structure format. The team created IARVs for several metrics mined from existing injury literature. The IARVs for the spine and chest deflection subsequently were refined based upon statistical models of combined non-injurious and injurious literature data. This work culminated in the development of an acceptable risk of injury from spaceflight, which included a newly developed operationally relevant injury scale.