A methodology is presented for determining distresses on asphalt and portland cement concrete pavements in an objective and reliable manner. Individual distress magnitudes are established using linguistic scales that consider distress severity and extent along nominal lengths of pavement. Emphasis is placed on generating both the quality and the quantity of data needed for pavement management purposes. The methodology is implemented in the form of a distress survey technique applied to the pavement system of the New York State Thruway. Data recording and handling are automated through the use of laptop computers. Yhe survey technique is taught annually to nontechnical personnel in a comprehensive training program. The results of statistical techniques used to evaluate rater performance and repeatability of survey data are presented and discussed. It is concluded that the implemented distress survey is a reliable procedure that produces distress data at acceptable levels of repeatability.