Moving Base Driving Simulators (MBDS) have a large potential to increase effectiveness in vehicle dynamics development. MBDS can reduce dependency on vehicle-prototypes by allowing subjective assessments (SA) of models. Little is, however, known about the relation of SA in MBDS and in physical vehicles. This paper aims to increase this knowledge, and proposes and implements a methodology to validate MBDS for SA of steering feel and handling. Firstly, vehicle models were generated from Kinematics & Compliance measurements of real vehicles. These models were validated versus objective tests, with steering robots, of the physical vehicles. These vehicles and their MBDS-models were assessed by expert drivers, using a scanned-test track in the MBDS. Comparison of the SA in both environments enabled the MBDS validation. Promising results, with higher SA accuracy for handling than for steering feel, indicates that the major improvement effort should focus on the steering model and its simulation in the MBDS.