Abstract This paper is primarily concerned with the skid behavior of tread compounds, and the extent to which testing parameters such as the type of road surface, the vehicle speed, or whether peak and sliding coefficients are considered, influence the skid behavior of one tread compound in relation to another. Whilst the actual braking coefficients depend profoundly on all these parameters the rating of two compounds is much less affected by them. Moreover, subject to the qualification given in the paper, the ranking of compounds will in many cases be predicted correctly by the RRL skid tester. The question arises how the rating of tread compounds will be affected by other testing methods. The three most commonly used are cornering, traction, and stopping distance tests. The first two are dominated by the peak value of the friction coefficient because the measurements are taken just before sliding occurs whilst the last reflects the sliding value of the braking coefficient. Since the rating is virtually the same for these two types of measurement it is likely to be independent of the measuring method. Although the ranking of compounds does not depend on the road surface, well drained road surfaces gave more reproducible results and larger ratings so that fewer readings are required on such a surface than on a flooded one. The following observations as to the nature of wet friction emerge from the present study. (a) Changes in coefficient of friction with tread compound and type of road surface texture confirm the findings of Miss Sabey, and Greenwood and Tabor, that the energy loss component of wet friction on coarse surfaces increases with the sharpness of the asperities, and thus with the pressure on the tops of the asperities. A mechanism for such energy losses based on elastic stored energy is suggested. (b) The decrease in the braking coefficient, as observed on well drained or dry road surfaces after the wheels have become locked, is explained by the frictional temperature rise in the area of contact. (c) It is suggested that in the case of the sliding tire, temperature increase in the contact region with increasing vehicle speed contributes to the observed fall in coefficient with increasing speed. This, together with the temperature dependence of the rubber properties, is used to explain small differences between the speed coefficients of various compounds in their rate of fall off with speed. (d) The mean braking coefficients of the compounds increase with their internal viscosity. It appears that the internal losses can be increased and therefore the skid resistance improved by incorporation of a heavy oil without an appreciable change in the glass transition temperature as measured by a torsion pendulum. This is strikingly demonstrated by the highly resilient natural rubber, whose skid resistance is greatly improved by oil extension without any significant loss in resistance to tire wear.