Accident data suggest that a significant percentage of rear impacts involve occupants seated in other than a “Normal Seated Position”. Pre-impact acceleration due to steering, braking or a prior frontal impact may cause the driver to move away from the seat back prior to impact. Nevertheless, virtually all crash testing is conducted with dummies in the optimum “Normal Dummy Seated Position”. A series of 7 rear impact sled tests, having a nominal AV of 21 mph, with Hybrid III dummies positioned in the “Normal Dummy Seated Position”, “Out of Position” and slightly “Out of Position” is presented. Tests were performed on yielding production Toyota and Mercedes Benz seats as well as on a much stiffer modified Ford Aerostar seat. Available Hybrid III upper and lower neck as well as torso instrumentation was used to analyze and compare injury potential for each set of test parameters. In all cases, neck forces and moments were found to increase when the dummy's torso was leaned forward at impact. For the out-of-position tests, the results showed that the upper neck loads were clearly related to seat stiffness; stiffer seats produced greater neck loading. The lower neck loads, however, showed no definite correlation with seat back stiffness.