The Intersection Decision Support (IDS) Project is designed to reduce crossing-path (CP) crashes at intersections by providing crucial information to drivers that would help them avoid such crashes. Over the past decade, researchers have used the General Estimates System (GES, a representative sample of police-reported crashes in the US) and other data sources to develop a taxonomy of CP crashes and pre-crash scenarios as groundwork for crash-prevention efforts. The current study builds on and extends prior work by constructing a taxonomy of CP crashes using data from the 2000 GES and identifying potential corresponding IDS countermeasures. Analyses differ from previously published analyses in that traffic control device data was available at the vehicle level, and not just at the crash level. This allowed more detailed study of crashes by traffic control device. Findings included documentation that crashes at intersections represent a very high percentage of all U.S. crashes, making intersections relatively high-risk areas compared to other roadway segments. Also, CP crashes constituted a substantial portion of total crashes in the US, including 25% of all crashes and about 45% of crashes at intersections. Patterns of CP crashes differed substantially by type of intersection (defined by traffic control device), and these differences in crash patterns reflected varied underlying causal factors that required tailored IDS countermeasures. In addition, CP collisions at intersections took place at moderate speeds, which is important for algorithms for warning systems. Finally, older drivers were over-represented in crossing-path collisions at intersections. IDS countermeasures will need to account for findings on intersections here and elsewhere that address driver behavior and vehicle movement and conflict.