The physical interaction between roadway surface and vehicle tires has been widely investigated. Roadway conditions, traffic characteristics, environmental factors, and driver habits all play important roles in skid resistance and safety characteristics of pavements. It is difficult, however, to use quantitative statements to describe these uncertain factors, and the poor correlation between pavement performance and these factors makes it impossible to predict skid-resistance properties with regression analysis. To evaluate pavement skid resistance, a new methodology has been introduced which uses fuzzy-sets mathematics. The skid performance of pavement sections can be grouped by fuzzy cluster analysis of the data obtained from sand-patch and British pendulum number testing, as well as skid numbers measured on test sections with blank and ribbed test tires. A computer model which predicts locations that have a high accident potential can be established from the fuzzy cluster results. These results can also be used to predict seasonal variations in pavement skid resistance. The principle discussed here can be valuable in all types of pavement-condition evaluations. This paper presents the state of the art in using fuzzy cluster analysis for evaluating pavement skid resistance, first describing the objectives and the scope of possible applications of this methodology, and then giving the calculation procedures and the results of a case study using this technique. Recommendations for future pavement-skid-resistance research are outlined.