Studies indicate that more people are exposed to noise from motor vehicles than any other source of noise. For this reason, the adverse noise impact of highway noise is of great concern to land use planners, regulatory agencies, and affected communities. To assess the impacts of new highway projects and to identify the effectiveness of mitigation measures, the federal government and some states require that noise analyses be conducted using specific computer models. STAMINA/OPTIMA is the Federal Highway Administration's approved model and is used by many states. SOUND 32 is the California Department of Transportation's preferred model, and is used for highway projects in California. Both models are based on similar mathematical formulations, but differ in vehicle noise emission characteristics, height of noise sources, and attenuation effectiveness for noise walls and earth berms. The purpose of this paper is to report the results of a comparative noise analysis in which both models were applied to an actual highway segment. The influence of traffic volumes, speeds, highway geometry, grade, and receptor distance on the model's estimated noise levels was studied. Differences in the location, dimensions, and effectiveness of barriers recommended by the two models are identified and some of the merits and drawbacks of the models are briefly discussed.