Abstract The dominant importance of hysteresis in lubricated friction of rubber was first outlined by Tabor and has been extensively confirmed since. A number of workers have also shown that dry friction is correlated with hysteresis, but the mechanistic connection is still unclear. Rubber hardness has been less definitely correlated with friction. This may result in part from the choice of surface on which to test the rubber compositions studied. We have previously shown the importance of the test surface and recommended use of wavy glass to simulate actual slippery roads in lieu of a representative road surface. Carr reported our finding that the relative influence of rubber hardness depends on the texture of the hard surface on which it slides. Using the wavy glass and a slippery road surface, we have now studied a very wide range of compositions, both gum and filled, derived from commonly used tire rubbers, as well as others included to extend the range of properties studied. These results fully confirm the importance of hysteresis and more clearly define that of hardness. These two properties together almost completely determine wet friction on slippery roads. Most of the variance not accounted for is attributable to measurement error.