The Virginia Department of Transportation (VDOT) Pavement Management Program has examined and improved the quality of condition data to carry out pavement performance analysis and multiyear work planning. In 1995 the agency changed the way it gathered pavement condition data, moving from subjective windshield surveys to using automated condition survey equipment. In 1997 the agency acquired the staff and resources to examine these data and respond to concerns of district personnel that the data were inconsistent and not representative of actual conditions. At that time, VDOT acknowledged the need for a formal and large-scale quality assurance program for its pavement condition data. It was shown in 1998 that standardization of test methods and calibration of equipment for roughness measurement yielded data of much higher quality than that in previous years. Quality data are crucial to the sound functioning of a pavement management system. It has been found that (a) distress data quality is a serious problem and the data can potentially be bad enough to be completely useless and (b) improving or reengineering distress data quality requires a significant effort. A structured approach to develop and implement a pavement distress data quality program at VDOT is described, including the processes, statistical details, and a clear vision of needs. Most important, this reengineering effort involves attending to the data collection process by building controls at critical junctures during the project in order to deliver a quality data product in time and on budget.