The growing applications of Computer-Aided Engineering (CAE) tools have been motivated by the need to create more effective product development processes. Computational Fluid Dynamics (CFD), as one of the CAE tools, has enjoyed growing popularity for analysis of many airflow situations, including road vehicle aerodynamics. In many cases, these applications have been limited by the level of predictive accuracy that is possible with CFD codes today. In the present exercise, simplified representations of three vehicle models (1:12 scale) were chosen to assess the overall level of predictability of the GMTEC CFD code, using detailed measurements that were made in a scale-model wind tunnel. The CFD computations used two turbulence models (standard k-εand RNG k-ε) and were matched to the experimental geometry and test conditions. The comparisons revealed that the CFD predictions were able to track relative changes in drag between the three models with acceptable engineering accuracy (less than 4 percent error of relative drag). Accurate prediction of the drag coefficient for each model was limited by many factors, including available computer resources, and the need for further understanding of the twin issues of turbulence modeling and steady-state simulation of inherently unsteady flows.