Objective The objectives of the study are to derive lower-neck-injury probability curves under rear impact loading from matched pair Hybrid III and THOR dummy tests. Methods Twelve whole-body and 15 isolated head-neck rear-impact sled tests were conducted using the 2 dummies. They were positioned on a rigid seat that was attached to an acceleration sled. The dummies were positioned with the head parallel to the ground, torso against the seat back, and legs stretched such that there was no axial rotation. The acceleration pulse matched the previous in-house human cadaver tests. The 6-axis lower-neck load cell was used in both dummies. For the isolated rear-impact tests, the head-neck was excised from the dummies, and the lower-neck load cell was mounted to the top of the sled with the head parallel to the ground and such that the acceleration vector was in the rear impact mode. Lower-neck loads and lower-neck-injury criteria (LNij) were obtained using the load cell data and survival analysis was used to develop injury-assessment-risk values and curves for both dummies. The LNij criteria were determined for both dummies using the intercept value corresponding to the 90% probability level for the forces and moments. Results The lognormal and Weibull distributions were the optimal distributions for the Hybrid III and THOR devices. At the 50% risk level, the mean LNij was 1.1 and NCIS was 0.77 for the Hybrid III, and 1.5 and 0.30 for the THOR device. The quality indices were in the fair range and good range for the 2 dummies, respectively, at this risk level. Conclusions The lower neck based LNij injury assessment risk curves and reference values, IARCs and IARVs, serve as the first dataset for injury assessments, and the THOR may be a better test device for assessing injures in rear impact environments.