Previous research on both small-scale and full-scale vehicles shows that base extensions are an effective method to increase the base pressure, enhancing pressure recovery and reducing the wake size. These extensions decrease drag at zero yaw, but show an even larger improvement at small yaw angles.In this paper, rear extensions are investigated on an SUV in the Volvo Cars Aerodynamic Wind Tunnel with focus on the wake flow and on the unsteady behavior of the surface pressures near the base perimeter. To increase the effect of the extensions on the wake flow, the investigated configurations have a closed upper- and lower grille (closed-cooling) and the underbody has been smoothed with additional panels.This paper aims to analyze differences in flow characteristics on the wake of an SUV at 0° and 2.5° yaw, caused by different sets of extensions attached to the base perimeter. Extensions with several lengths are investigated with and without a kick. Unsteady pressure sensors attached to the base perimeter of the vehicle are used. The fluctuations are examined for the reference vehicle and the vehicle with extensions for different freestream velocities and under a yaw angle ranging from 0° to 7.5° yaw.The current investigation indicates that the drag reduction obtained at 2.5° yaw with a “kick” mounted at the rear edge of the extensions have an effect on the size of the wake and enhance the pressure recovery behind the vehicle. The addition of the kick alters the wake distribution and changes the corresponding flow pattern with the largest effect under yaw conditions.For the unsteady surface pressure measurements, a technique based on Empirical Mode Decomposition (EMD) is applied that results in instantaneous frequencies of the signal. It gives the option of signal reconstruction limited to the frequency area of interest. A Proper Orthogonal Decomposition (POD) on the signals is also conducted, showing large variance in the area below the catwalk for the first two modes. To enhance the correlation, this paper investigates the combination of both mode decompositions, where the POD is applied to EMD-filtered signals.