Injury criteria are crucial in whiplash protection evaluations. Therefore, the real-life rear impact performance of eight car seats was compared with various injury criteria using linear correlation techniques. Two dummies, BioRID and RID 3D, and two types of pulses were used: generic and car specific. This evaluation showed an acceptable correlation of the lower neck shear force measured in the RID 3D dummy with the real accident data. A reasonable amount of correlation was found also for the NIC measured in BioRID with a car specific pulse. When the injury risk figures were compensated for real-life car exposure, no correlations were found for any dummy using the generic pulse.