Cross median crashes have become a serious problem in recent years. Most of the median cross sections used for divided highways have terrains with steep slopes. Traffic barriers, frequently used on slopes, are generally designed based on the findings obtained from crash tests performed on flat terrain. For barriers placed on roadside and median slopes, vehicle impact height varies depending on the trajectory of the vehicle along the ditch section and lateral offset of the barrier. Thus depending on the placement location on a relatively steep slope, a barrier can be impacted by an errant vehicle at height and orientation more critical compared to those considered during its design. Hence, detailed study of performance of barriers on roadside and median slopes is needed to achieve acceptable safety performance. In this study, performances of modified G4(1S) W-beam, Midwest Guardrail System (MGS), modified Thrie-beam, modified weak post W-beam, and box-beam guardrail systems on sloped terrains are investigated using numerical simulations. A procedure is developed that provide guidance for their placement on roadside and median slopes. The research approach consists of nonlinear finite element analyses and multi-rigid-body dynamic analyses approach. Detailed finite element representation for each of the barriers is developed using LS-DYNA. Model fidelity is assessed through comparison of simulated and measured responses reported in full scale crash test studies conducted on flat terrain. LS-DYNA simulations of vehicle impacts on barriers placed on flat terrain at different impact heights are performed to identify performance limits of the barriers in terms of acceptable vehicle impact heights. The performances of the barriers are evaluated following the guidelines provided in NCHRP Report 350. Multi-rigid-body dynamic analysis code, CARSIM, is used to identify trajectories of the vehicles traversing various roadside and median cross-slopes. After analyzing vehicle trajectories and barrier performance limits, a guideline has been prepared with recommendations for the placement of barriers along roadside and median slopes. This guideline is then verified and refined using the responses obtained from full-scale LS-DYNA simulations. These simulations capture the full encroachment event from departure of the vehicle off the traveled way through impact with the barrier.