The National Highway Traffic Safety Administration (NHTSA) estimates delta-V from detailed measurements of vehicle deformation using the WinSMASH crash reconstruction code. Previous research has shown that WinSMASH delta-V estimates underpredict true delta-V by 25% on average. One possible explanation for this error is inaccuracies in the stiffness values used in the delta- V reconstruction calculation. The accuracy of codes, such as WinSMASH, is dependent upon vehicle stiffness values computed from post-impact crush measurements in crash tests. Any error in these crush measurements will be reflected as inaccuracies in the stiffness coefficients, and ultimately as errors in WinSMASH delta-V estimates. This paper investigates the accuracy of post-impact crush measurements in 93 frontal New Car Assessment Program (NCAP) tests of model year 2005-2007 vehicles.