This study presents a methodology based on fuzzy set theory that enables the determination of a single Pavement Distress Index (PDI) from commonly used surface distresses. The magnitudes of individual distresses are established using linguistic scales. These have been formulated on the basis of pattern classification techniques that consider both the severity and the extent of each distress along nominal lengths of pavement structures. Linguistic distress values are represented by membership functions and are quantified using input from maintenance experts. A single measure of overall pavement damage is determined through fuzzy weighted average computations and linguistic approximation. The methodology is illustrated in a case study, and important conclusions are presented and discussed.