Several researchers have raised anecdotal concerns that NASS/CDS may overestimate ΔV in side crashes. NASS/CDS investigators use the WinSmash code, a successor to CRASH3, to perform the estimations The objective of this study was to determine the accuracy of WinSmash reconstruction of ΔV in side crash tests. The actual ΔV and absorbed crash energy were computed for a suite of 73 NHTSA side crash tests using crash test instrumentation. Multiple accelerometers on both the striking and struck vehicle were used to calculate full planar motion histories, vehicle rotation, and centerof- gravity ΔV at maximum crush and at vehicle separation. The same crash tests were then reconstructed using WinSmash and post-test crush measurements. This paper compares the WinSmash ΔV with the actual ΔV at maximum crush and ΔV at separation. The paper concludes that WinSmash over-predicts ΔV at separation in side crash tests by 11% on average.