The accuracy of tyre models depends to a large extent on the measurement data used to assess model parameters. The MF-Swift tyre model parameters can be identified or estimated from various combinations of experimental data. The amount and required accuracy of the measurement data can be selected according to the tyre model application. This paper discusses a simulation study carried out using TNO's MF-Swift implementation in Simulink (see www.delft-tyre.com) using sets of parameters obtained from different combinations of measurement data that have been supplied for the tyre model performance test benchmark. The results show the relation between measurement effort and model accuracy, thus relating the most efficient test protocol to the tyre model application.