As part of a FHWA sponsored research project to study highway system resilience, a two-fifths scale curved steel plate girder bridge was constructed and subjected to earthquake simulation at the Large Scale Structures Laboratory at the University of Nevada, Reno (UNR). The objective of this simulation was to study the seismic system behavior of the bridge as well as additional components including reinforced concrete columns, effects of live load, isolation systems, ductile-cross frames, and abutment behavior. Ultimately design recommendations will be developed from this research. The research that is presented in this document is the results of the design, analysis, and experimental results of the conventional bridge columns and substructure elements as part of the research being conducted at UNR. The design of the substructure elements was completed according to the requirements of the AASHTO Guide Specifications for LRFD Seismic Bridge Design. In addition, the column design was based on the typical column sizes used by the local departments of transportation. The Sylmar recording of the 1994 Northridge, California earthquake was used as the input ground motion in the system. Analytical modeling using SAP2000 was performed on the scaled bridge model to estimate the seismic response of the bridge using non-linear time-history analysis. Numerical analysis was used to check the system at the design level earthquake and at a large amplitude motion intended to cause column failure. In addition, the analytical models were subjected to the testing protocol, ten ground motions with increasing amplitudes, to determine the effect of the loading protocol on the system. The response of the columns during experimental testing met all performance requirements at the design level and maximum considered earthquakes. The effects of shear keys in the system were shown to have an impact on the torsional loads in the system. At the end of the last test, longitudinal reinforcement started buckling in the columns, however; columns had not reached the maximum lateral capacity. Testing was stopped at this point due to shake table limitations.