In 2000 a three-year project named Tyre and Road Wear and Slip assessment (TROWS) was started. One of the TROWS objectives was to provide a tool able to numerically predict tyre global wear as well as to qualitatively determine the wear distribution. The proposed methodology combines a mathematical model of the tyre with an experimentally determined local friction and wear law. Thus, tyre abrasion due to each single manoeuvre can be determined. The tyre model (called PaRRT model, Physical Rigid Ring Tyre model) allows the simulation of any given manoeuvre starting from the time histories of the applied generalized hub forces, i.e., the longitudinal, lateral and vertical hub forces, the applied driving/braking torque and the imposed slip and camber angles. These input quantities may come from experimental measurements or from vehicle's multi-body simulations. The tyre structural model is made up of three rigid bodies (a disc, a ring and a plate) that represent the hub, the belts and the contact area between tyre and road. Full-scale experimental tests were carried out with two Peugeot 406 cars on a public road course in Italy. Each car was equipped with a different set of tyres: one car was equipped with four all-season tyres (called A tyres) and the other car was equipped with four winter tyres (called B tyres). Both sets of tyres had a 195/65 R15 size. The collected data was used to validate the model. The validation of the model was carried out in terms of total wear depth rate. Great attention has to be paid to the tyre contact grid. Otherwise, the tyre wear will probably be underestimated as is the case of B tyre. Moreover, in order to correctly determine tyre wear over a long track, several iterations of vehicle model and tyre model simulations are necessary. Moreover, a numerical procedure to investigate the effect of tyre design parameters on tyre wear was set up. A series of 'virtual' tyres was generated by varying each single parameter of the physical tyre model by 25%. The virtual tyres wear behaviour was compared with the wear behaviour of a reference tyre on an ideal track using a segment C car model driven by a very simple driver model. As expected, the sensitivity analysis showed that the tread compound properties are the ones that mostly affect the slip distribution inside the contact footprint and thus the wear behaviour of the tyre. An increase in tread compound density, stiffness and damping would lead to a sensible reduction of tyre wear (respectively, around 20%, 30%, 15%). It should however by taken into account that these improvements are achievable only by keeping friction and wear properties of the tyre constant. Although the 'best tyre' obtained from this work may not be obtainable with the present technological solutions, this analysis provides a quantitative indication of which of the tyre parameter is most effective to reduce tyre wear.