owards the development of a dummy-based thoracic injury risk function, previous studies have emphasized the need to determine an injury predictor variable derived from dummy deflection measurements such that the injury prediction is independent of the type of restraint loading. The goal of this study was to evaluate the significance of multi-point deflections measured on the anterior and lateral rib cage in predicting the risk of rib injury in different frontal crash loading scenarios. The methodology of the study included statistical analyses using regression techniques to correlate the risk of severe rib injury and maximum strain in the ribcage with linear combinations of multi-point deflections as explanatory predictors. Thoracic multi-point deflection data used for regression analyses included data from 20 sled tests involving belt restrained Post Mortem Test Subjects (PMHS) and simulation based thoracic deformation data involving two finite element human models oriented in 18 different loading conditions. The loading conditions for the simulation database included table top loading with 3 point belt, 4 point belt, airbag and hub restraints; static airbag and pendulum loading for seated occupant; and dynamic sled tests with belt only, airbag only and combined belt and airbag restraints. In addition to the evaluation of the deflection predictors, qualitative predictors, describing the loading environment, were evaluated including the presence of belt and airbag restraint, hub impact, and load-limiting restraining force. The results demonstrated that for frontal loading of the thorax by the 3 point belt system, the mid- sternum deflection and existing Cmax measure performed better compared to the composite multi- point deflections in predicting the risk of severe rib injury in PMHS. The simulation results which characterized the thoracic deformation in different restraint loading conditions, loading rates and two different human models, suggested that lateral deflections in addition to the standard four point anterior deflections were significant to the prediction of the maximum strain in the rib cage. The linear combination of multi-point deflection predictors as derived in this study may be further modified to provide the framework for dummy-based injury risk function applicable to a range of standard restraint conditions in frontal crashes.