Petit et al. 2015 and Lebarbé et al. 2016 reported on two studies where the injury mechanism and threshold of the sacroiliac joint were investigated in two slightly oblique crash test conditions from 18 Post Mortem Human Subjects (PMHS) tests. They concluded that the sacroiliac joint fractures were associated with pubic rami fractures. These latter being reported to occur first in the time history. Therefore it was recommended not to define a criterion specific for the sacroiliac joint.In 2012, injury risk curves were published for the WorldSID dummy by Petitjean et al. For the pelvis, dummy and PMHS paired tests from six configurations were used (n = 55). All of these configurations were pure lateral impacts. In addition, the sacroiliac joint and femur neck loads were not recorded, and the dummy used was the first production version (WorldSID revision 1). Since that time, the WorldSID was updated several times, including changes in the pelvis area.In the present study, the two slightly oblique PMHS test configurations of 2015 and 2016 (n = 18) as well as three of the reference configurations used in the 2012 IRC (n = 43) plus one impactor configuration from WSU (n = 9) were duplicated using the 50th percentile male WorldSID dummy build level F. In these tests, the dummy was systematically equipped with the SI-joint and femur neck load cells.The results show that the sacroiliac Fy is a better pelvic ring injury predictor than the pubic load. The aim of the paper is to provide an analysis of the paired test results and to propose both an updated pelvic injury criterion and the injury risk curve (IRC).