The THOR-AV prototype was designed as an immediate tool for the industry to test the restraint systems for autonomous driving system (ADS) equipped vehicles. In this study, the THOR-AV was evaluated for its responses in the traditional front seat and rear seat configurations. The National Highway Traffic Safety Administration’s (NHTSA) latest BioRank method was used to objectively assess the THOR-AV biofidelity. The interaction between THOR-AV and the restraint system in the front and rear seat configurations had BioRank scores (BRS) of 1.54 and 1.64 respectively, both corresponding to good biofidelity. THOR-AV responses had BRS scores of 0.84 and 0.77 for the front seat and rear seat respectively, both corresponding to excellent biofidelity. Overall, the THOR-AV interaction with the restraint system had a combined BRS score of 1.59, corresponding to good biofidelity. The THOR-AV responses had a front and rear seat combined BRS score of 0.80, corresponding to excellent biofidelity. It was observed that the THOR-AV prototype submarined slightly earlier than post-mortem human subjects. Design changes were explored using the THOR-AV finite element model. The revised design improved the pelvic bone geometry and pelvis flesh compression, the dummy pelvis kinematics and submarining time.