The aim of this study was to evaluate the biofidelity of the CAVEMAN human body model lumbar spine and to evaluate the effects of variation in lumbar spine posture on load transmission and injury location. The CAVEMAN model was validated against vertical impact experiments at sub-injurious loading rates using data sets derived from post-mortem human subject testing. The spine model was positioned to represent the curvature of six post-mortem human subject spines, and each posture model was then loaded vertically with identical input conditions. These simulations were then repeated at an injurious loading rate, and the predicted fracture locations and severity were compared. The load response of the CAVEMAN spine models were found to be within the measured range of the post-mortem human subject response and within ±1 standard deviation of the mean response through the time of peak impact. The magnitude of peak load measured in the models, as well as predicted fracture locations and severities, were dependent on spine posture, with sacrum angle the predominant factor.