Frequent measurements of skid resistance were made on 20 pavements in common use in Kentucky. Tests were conducted from November 1969 through 1973. Principal analysis involved relating changes in skid resistance to day of the year and relating skid resistance to temperature at the time of test, to average antecedent temperatures, and to average rainfall.

Seasonal variations exhibited an annual sinusoidal cycle. The changes in sand-asphalt and bituminous concrete surfaces under higher volumes of traffic were about 12 skid numbers. The changes in portland cement concrete and bituminous concrete under lower volumes of traffic were about 5 skid numbers. The lowest skid numbers (SN's) occurred in early to mid-August for portland cement concrete and sand-asphalt pavements and in late August to early September for bituminous concrete. Correlations between changes in SN and temperature were best for ambient air temperature averaged over 4-and 8-week periods prior to date of test. However, correlations between changes in SN and temperature were not as good as correlations between SN and day of the year. On the other hand, combining traffic volumes in the form of deviations from yearly average daily traffic with temperature yielded correlations with SN's which were as good as correlations between SN's and the day of the year.