toring sources of sodium in large portions of the food supply and can be scaled based on the availability of resources for conducting the analysis. Once key targets are identified and the methodology is established, this method could provide data on trends over time by individual food categories. However, the analysis will have some limitations that may have to be addressed using other sources. In particular, private label products that are contract-manufactured for the major retail chains are included in the store scanner data, but it may be infeasible to link these foods to nutrient data. The analysis will likely need to focus on the top-selling products based on the availability of nutrient data at the UPC level from outside vendors. Foods that are prepared and packaged within a retail establishment are not currently required to include the Nutrition Facts panel; thus, the sodium content of these foods cannot be monitored without linking through other types of data sources that would require substantially more manual effort. Not all stores participate in store scanner data reporting (e.g., Wal-Mart), but product sales from these stores can be tracked using household-based scanner data because at least a portion of the household panel purchases items at stores not currently captured in store scanner data.