The talk will discuses an appropriate usage of large eddy simulation (LES) in external vehicle aerodynamics. Similar as for other tools used in vehicle aerodynamics, LES application must be chosen with greatest care to avoid inappropriate usage. Flows in which experimental investigation are difficult to perform due to either physical limitations or practical problems to duplicate operational flow conditions (e.g. boundary conditions) in an experimental set-up are appropriate target of LES in vehicle aerodynamics. Three different applications including wheelhouse flow, gusty flow and active flow control, are used to demonstrate how LES can be used to obtain new knowledge about vehicle flows. The example of the wheelhouse flow is representative of flows where experimental investigation has limitations due to physical constrains of experimental equipment and where LES is an excellent choice of numerical technique due to the separated character of the flow. The flow around the oscillating model illustrates how LES can increase our knowledge of flows where the movement of the vehicle or transient character of the upstream flow is critical for accurate representation of the flow and vehicle's performance. The potential of LES usage for this kind of applications is large due to its importance for crosswind stability and comfort of vehicles. The last application of the active flow control gives an idea of future usage of LES for improvement of aerodynamic performance of vehicles. Results of three different flow control situations will be presented. The first two cases use periodic blowing and suction to control development of wake behind 2D and 3D Ahmed body [1]. The third case uses constant flowing or suction to control development of longitudinal vortex that develops on the A-pillar from the experimental investigation in [2]. The usage of active flow control will win ground in many vehicle applications and its understanding and optimization of flow control parameters can only be done using experimental techniques or flow resolving numerical techniques. The latter in form of LES will in the future for sure show to be more cost efficient. References: 1. Krajnovic S and Fernandes J., Numerical simulation of the flow around a simplified vehicle model with active flow control, International Journal of Heat and Fluid Flow, 2010, doi:10.1016/j.ijheatfluidflow.2010.06.007 2. Lehugeur B., Gillieron P. and Kourta A., Experimental investigation on longitudinal vortex control over a dihedral bluff body, Exp. Fluids (2010) 48:33-48