Underrun Protection devices (UPDs) are specially designed barriers fitted to the front, side, or rear of heavy trucks. In case of accidents, these devices prevent small vehicles such as bikes and passenger cars going underneath and thus minimizing the severity of such accident. Design and strength of UPD is such that it absorbs the impact energy and offers impact resistance to avoid the vehicle under run. Compliance to UPD safety regulations provides stringent requirements in terms of device design, dimensions, and its behavior under impact loading. Since accuracy of Computer Aided Engineering (CAE) predictions have improved, numerical tools like Finite element method (FEM) are extensively used for design, development, optimization, and performance verification with respect to target regulatory performance requirements. For improved accuracy of performance prediction through FEA, correct FE representation of sub-systems is very important. One such sub-system in UPDs is bolted connection. Modeling of bolted connection requires correct representation of critical attributes of bolted joint, to accurately capture bolted joint behaviour under extreme loading situations. The two primary important bolted joint attributes are bolt pretension and a mating part contact. This paper compares the UPD FE analysis with two types of bolt modelling methods, a) solid elements method b) a combination of beam, shell and constrained spot-weld elements method. The solid element method can accurately model mating part contact behavior of the structure. In this case material stiffness representation for solid elements is based on an elasto-plastic material, defined by a piecewise stress-strain curve and a strain rate dependency. The bolted joint failure is modeled through strain-based failure criteria. With this method it is being observed that, pretension modeling is very tedious and it results into numerical instability. In second method mating part contact is achieved through shell elements modelling and failure is based on the “constrained spot-weld failure”. Failure definition requires failure load values of bolted joint in axial & shear loading condition. These failure load values are usually derived from component level physical tests on bolted joint. In this study, bolted joint performance prediction for RUPD with both modeling methodologies is compared with physical test results.