unit prices for each item of work on every project. There are several ways to estimate: resource cost, reference, and historical cost. Each has its advantages and disadvantages and carries its own degree of uncertainty. Resource cost–based estimating is the most accurate and reliable method. It depends on the precision of the direct costs, applied indirect costs, and certainty of the best information available at the time. However, it is impractical for highway departments since it is time intensive and impossible to know each contractor’s crew configurations, crew costs, production rates, equipment status, markup costs, and so forth. Therefore, highway departments must use other means to develop reliable cost estimates. State highway departments typically estimate construction proj- ect costs based on historical bid data (2). With over 100 bids being let each year in Arizona, a tremendous amount of historical data is generated. All the information has the potential to be vital. However, not all the stored information is needed to estimate items of work on future projects. Only relevant or project-specific information is use- ful. Factors must be identified that are likely to have the greatest affect on item unit prices on a project. Primary variables that cause unit prices to fluctuate are quantity, geographic location, and project type and characteristics. The preparation of a project cost estimate requires the collection, storage, retrieval, and manipulation of tremendous amounts of infor- mation in a timely manner. A systematic process needs to be employed that is capable of retrieving pertinent data more quickly more easily, and less expensively. Previous projects can then be searched to identify similar or identical work activities. It is difficult to extract and analyze cost information statewide when information is not together. A great deal of time is required to page through and assimilate reams of compiled data. Additional delays occur in determining whether the data are in the desired geographic location. Valuable time is expended by referencing back and forth. Analysis would be greatly simplified if the data could be presented graphically. A quick visual observation could identify data of interest and thus facilitate decisions. Queries on one or more fields could be added to further streamline the records and capture the pertinent information more precisely. Quicker, more informed decisions regarding unit prices could then be made, and the certainty, accuracy, and cost-effectiveness of estimates would increase. This research is the first to use GIS in construction cost estima- tion. An extensive literature review was conducted to locate and identify other research efforts on the use of GIS in cost estimation. Numerous articles on GIS applications in traditional areas of a high- way agency’s responsibility were found: planning and research, traf- fic engineering, pavement management, bridge management, and routing. Many current as well as proposed applications of GIS are Geographic Information Systems as a Support Tool in Construction Cost Estimating in State DOTs SULEIMAN A. ASHUR AND BARRY CROCKETT 112 Paper No. 971231 TRANSPORTATION RESEARCH RECORD 1575 S. A. Ashur, Department of Civil Engineering, University of Texas at El Paso, El Paso, Tex. 79968-0516. B. Crockett, Contracts and Specifications Section, Arizona Department of Transportation, 1651 W. Jackson, 121F, Phoenix, Ariz. 85007-3276. Geographic Information Systems (GIS) have been successfully imple- mented in various fields. Transportation agencies, however, have not yet recognized the full capabilities of GIS. The purpose of the GIS cost estimation project is to research the possibilities of analyzing data through the power of geographic management. The enhanced data management would assist in improving project cost estimation. Typi- cally, state highway departments rely heavily on historical bid data to estimate construction projects. A systematic information collection, organization, and storage process must be used so that relevant histor- ical cost data can be retrieved. Data collection and storage have been done for years but their usefulness at times appears worthless, primar- ily because of the amount of time required to page through and assim- ilate reams of compiled data. Analysis would be greatly simplified if one could visualize the data graphically. Using such technology should assist in easing the ever-increasing demand to analyze information to support more effective decision making. State highway departments estimate the cost of hundreds of con- struction projects each year. Estimating sections typically rely on historical bid data to estimate construction project costs. Data col- lection and storage have occurred for years but often appear under- used, primarily because of the time required to page through and assimilate reams of compiled data. Analysis would be greatly sim- plified if one could quickly retrieve and visualize the data graphi- cally. Decisions and conclusions could then be made intuitively on pertinent data, thereby rapidly identifying records of interest. The field could then be further narrowed through additional refinement. Technological advances in hardware and software coupled with declining costs have opened the door for using geographic informa- tion system (GIS) databases in this manner. Such technology should assist in easing the ever-increasing demand to analyze information to support more effective decision making. GIS have been applied in many fields outside the transportation industry. However, the full capabilities of GIS for transportation have yet to be realized (1). The purpose of the GIS cost estimation project is to research the possibilities of analyzing data through the power of geographic management. Enhanced data management would assist in improving cost estimation of transportation projects. BACKGROUND Each year, more than 100 contracts are let by the Arizona Depart- ment of Transportation (ADOT). The engineering staff estimates