Traditional vehicles are designed to be very stable and ensure no loss of control by the driver. High stability is formally translated into a large positive stability margin (SM) which leads to highly understeering vehicles. However, a increase in stability often causes a decrease in peak lateral grip. On the other hand, a lower SM can cause a greater phase lag in the vehicle’s frequency response which makes the control of a vehicle at limit handling conditions very complicated. With autonomous vehicles, the human factor is eliminated and, consequently, so is the need for high stability. Therefore, it could be possible to exploit the passive vehicle dynamics and enhance the performance, both in terms of peak grip and phase lag response. The goal of this paper is to explore the dynamic characteristics of a vehicle by comparing results of a vehicle with different levels of stability when driven by a human and an autonomous “driver” on a non-standard double lane change manoeuvre. This has been done by considering different configurations of the same vehicle which give different values of SM and peak lateral acceleration. After having analysed the dynamic response of the different configurations in both steady and unsteady state, several tests were run on a dynamic driving simulator (DiM) driven by different human drivers. The same tests were then run again in a model-in-the-loop (MiL) simulation where the vehicle is controlled by means of a model predictive control (MPC). The results show that the robotic controller outperforms a human driver and poses interesting design challenges for the vehicle with the advent of autonomous vehicles due to the increase in delays and instabilities.