Sled tests were conducted at a velocity of 6.7 m/s using side impact dummies (SID, BioSID, ES-2, and WorldSIDp) and the resulting biomechanical responses were compared with responses from post mortem human subjects (PMHS). Initial impact conditions were with and without pelvic offset in combination with and without padding on the impacting wall. Impact forces, thoracic trauma index, chest compression, and viscous criteria were evaluated. The probability of injury was estimated and rates of deformation were computed for each body region. Dummy responses were not always similar in terms of trend and injury criteria when compared with PMHS tests under the same initial conditions. Response variations will be of value in improving the biofidelity characteristics of dummies for crashworthiness evaluations.