In the 54 years since passage of the National Historic Preservation Act, more than 56 million ha of land have been surveyed in the United States, and nearly one million cultural resources have been identified and recorded. These efforts have produced hundreds of thousands of project reports, vast collections of data, and a wealth of descriptive information about the past. The accumulated data can be used to generate important new knowledge about the past, with many scientific and management implications, but remain largely untapped. Following current approaches, many resources will be damaged or lost before effective strategies for studying or preserving them can be developed. Synthesis and modeling are needed in creative mitigation efforts to identify which resources to preserve and study and how best to do so with limited time and funding. This article explores the potential for compiling and synthesizing large cultural and environmental datasets within a geographic information system to model the nature and distribution of cultural resources. It is argued that dedicated synthesis and modeling of cultural resource management data will allow development of more effective and proactive research and management strategies, providing lasting benefit to diverse scientific and traditional communities and the public.