This paper explores some methods for circumventing the difficulties in interpretation of skid and traction tests caused by variations in speed, water depth, and road surface.Testing data are analyzed with a new approach using a graph of peak coefficient against slide coefficient. These analyses show that the tire and surface do not change greatly in basic frictional response as test speed changes. However, changes in coefficient do occur on most wet surfaces as the test speed is varied. Data indicate this to be the result of changes in the percentage of the tire penetrating the water film to reach the road surface.Data obtained under nearly identical conditions on different days vary. Least variation occurs in the low-speed peak coefficients, the slide coefficients, and the sum of these coefficients obtained either on dry surfaces or on surfaces wet with very thin water films.The vector line recording the change of the wet peak and slide coefficients as they vary with speed is useful in determining the quantity of water depth actually present on a surface.Methods of using these data are suggested which reduce the testing variation under all conditions, help to characterize surfaces, adjust data for day-to-day test variation in water depth, and interpret the data in a more meaningful manner.