For generations, highway projects in the United States have been delivered with the traditional design-bid-build (DBB) project delivery method. In 1990, FHWA introduced Special Experimental Project Number 14 (SEP-14) to encourage state agencies to use more efficient alternative methods for delivering projects. As of 2002, approximately 140 SEP-14 projects have been built according to the design-build (DB) delivery method. Although the use of the DB method for transportation projects has increased noticeably in the past decade, attempts to measure project performance under DB as against DBB have been limited to small projects. To fill this research gap, an input-versus-output benchmarking approach was used to assess the performance of large DB highway projects. By means of this approach, factors that could affect performance of highway projects were identified, and a comprehensive methodology for benchmarking was developed. A sample of four DB projects (with contract amounts ranging from $126 million to $1.4 billion) was analyzed to determine how project characteristics and DB practices (inputs) affect project performance in terms of cost and schedule (outputs). Project cost and schedule performance show clear trends against 15 input factors. More data points are needed to determine correlations between these inputs and outputs. In addition, 11 DBB projects (with contract amounts ranging from $50 million to $100 million) were selected for a statistical comparison with DB projects.