Recent regulations on greenhouse gas (GHG) emission standards for heavy duty vehicles have prompted government agencies to standardize procedures to assess aerodynamic performance of Class 8 tractor-trailers. The coastdown test procedure is the primary reference method to assess vehicle drag and other valid alternatives include wind tunnel testing and computational fluid dynamics (CFD) simulations.While there have been many published studies comparing results between simulations and wind tunnel testing, it is less well understood how to compare results with coastdown testing. Both the wind tunnel and simulation directly measure aerodynamic drag forces in controlled conditions, while coastdown testing is conducted in an open road environment, aerodynamic forces are calculated from a road load equation, and variable wind and vehicle speed introduce additional complexity.In this paper we evaluate different methods to improve the correlation of absolute drag values between simulation and coastdown testing. Coastdown tests were performed on a 2010 ProStar Long Sleeper with a standard 53-foot dry-box trailer at Ford Proving Ground in Arizona per EPA's modified J1263 & J2263 procedure. The CFD simulations use a Lattice Boltzmann based method on a high fidelity geometric model. The first drag results are established by using validated best-practice CFD methodology. Subsequently, we improve the comparison to coastdown test results by accounting for the aerodynamic torque on the wheels, the detailed shape of the tires, and the impact of different surface roughness values on the tractor-trailer surfaces. Finally, we discuss the impact of the variable vehicle speed and wind conditions during the testing.