Abdominal injuries represent a small proportion of crash injuries but it increases considerably with regard to serious to severe injuries. It is important to assess abdomen injury risk using a biofidelic crash test dummy, especially under seatbelt loading. A lumped element model has been used to simulate seatbelt and impactor loading cases on Post Mortem Human Subjects (PMHS) subject and on the THOR dummy. The model consists of two blocks of one spring and one damper with a mass between the blocks and a mass at the front of the model. The front block has the spring and damper in series, whereas they are in parallel for the rear block, to reproduce the abdomen and back force respectively. The mechanical parameters of the model have been optimised to fit the force‐penetration response of PMHS and of the dummy. The results show that the PMHS response is more damper‐dependent, whereas the dummy response is more spring‐dependent. Moreover, unlike the PMHS, the THOR dummy model needs a non‐linear spring in order to fit the impactor test data. The conclusions are that the dummy abdomen should be modified to make it more viscously deformable and to include pelvis design modifications in order to prevent a too stiff response