Seventeen side impact sled tests were performed using a horizontally accelerated sled and a Heidelberg-type seat fixture. In these tests the subject's whole body impacted a sidewall with one of three surface conditions:​ 1) a flat, rigid side wall, 2) an unpadded side wall with a 6″ pelvic offset, or 3) a flat, padded side wall. Forces, deformations and accelerations of the thorax were measured. VCmax emerged as a good injury criterion in this test series. In the WSU-CDC tests VCmax was 0.86 m/s for the half-chest at 25% probability of AIS 4. In addition, a new criterion we have termed “Average Spine Acceleration” (ASA) performed well. ASA is obtained by integrating the T12-y spinal acceleration to obtain spine velocity. The slope of the spine velocity trace is the Average Spine Acceleration, and reflects the rate of momentum transfer to the body during an impact. ASA performed better than peak spine acceleration in three sets of cadaver data bases:​ WSU-CDC sled tests, GM-WSU pendulum impacts, and NHTSA-sponsored sled tests. In a Weibull analysis of combined NHTSA-WSU tests values of 25% probability of AIS 4 were as follows:​ TTI 145 g's and ASA 30 g's (age and mass normalized).