Injuries to the lower extremities are one of the major issues in vehicle-to-pedestrian collisions. To evaluate pedestrian lower extremity protection, the Transport Research Laboratory (TRL) legform impact tests have been conducted according to the specifications in the EU directive. The TRL legform impactor consists of a tibia and a femur steel shaft connected by deformable knee bars. A Flexible Pedestrian Leg-form Impactor (Flex-PLI), which has flexible femur and tibia, is examined in the Global Technical Regulation (GTR). Previous studies compared the responses of both impactors; however, the relation between the tibia acceleration in the TRL legform impactor and the maximum bending moment in the Flex-PLI (both injury measures are for the tibia fracture) is not understood sufficiently. In this study, using finite element (FE) analyses, the injury measures of the TRL legform and the Flex-PLI were compared with those of the human FE model in impact with a simple large car model to identify which injury measures of the two impactors are reasonable to predict injury risk to pedestrian lower extremity.The rigid body model was used to gain a theoretical understanding of the injury measures of the impactors. Based on the equations of motion of the rigid body model, physical parameters were identified that relate with the injury measures. The tibia acceleration relates with the bumper force and the spoiler force. The bending moment depends on the equation that is expressed by the force and distance of the bumper and the spoiler energy absorber. Therefore, the tibia acceleration and the bending moment are different design parameters. The knee bending angle depends on the moment around the tibia center of gravity. Analyses of the TRL legform, Flex-PLI and human FE models were conducted; and it was shown that the injury measures in the FE models were affected significantly by the physical parameters determined in the rigid body model.The responses of the TRL legform, the Flex-PLI, and human FE models were compared with changing vehicle parameters. The trend was comparable for the three models. However, when the bumper energy absorber was soft and its bottoming-out occurred, the tibia acceleration in the TRL legform increased sharply, whereas the bending moment in the Flex-PLI did not increase. In the human FE model, when the bumper energy absorber bottomed out, the stress of the tibia that made contact with the bumper energy absorber was still small. As far as changing parameters of the large car in the present study, the responses of the Flex-PLI are more comparable with the human FE model than those of the TRL legform impactor; and it is likely that the Flex-PLI can predict injury risk more reasonably than the TRL legform impactor.