The study presented herein contributes to the continual evolution of seismic design of geofoam roadway embankments by providing a comparison of the 1 % and 2 % strain criteria for seismic load-bearing analysis and providing a preliminary methodology to estimate the compressive resistance at 2 % strain because ASTM D6817 does not currently provide compressive resistance values at 2 % strain. Additionally, this study presents a methodology for evaluating static and dynamic stresses from numerical analysis results for use in selecting the appropriate geofoam types to include within the embankment and one approach of evaluating numerical horizontal sliding results as part of the overall seismic internal stability design process. These new seismic analysis methodologies are demonstrated by analysis of one of the geofoam embankments that the Tennessee Department of Transportation (TDOT) is proposing as partial replacement of an existing 478 m length bridge along Poplar Avenue in Memphis, Tennessee located in the mid-southern part of the United States. The 2 % criterion for seismic load-bearing analysis provides an economical advantage compared to the 1 % criterion because the 1 % criterion will typically require the use of higher density geofoam blocks and the cost of geofoam blocks increase with density. Although the 2 % axial strain criterion is currently based on limited test data, if static and temporary dynamic stresses do exceed the cyclic elastic limit of the EPS, the potential consequence can be that long-term creep strains after the earthquake event may be greater than anticipated resulting in overall settlement of the embankment that can be greater than anticipated. However, the additional deformation and settlement of the embankment would typically not result in catastrophic collapse, which is in general agreement with the ‘no-collapse’ philosophy for seismic design that has been adopted by many state and federal agencies.