Americans spend about half of their food budgets to purchase about two-thirds of their food from stores. USDA purchases retail and household scanner data for food and economic research covering a broad range of Federal food and nutrition topics that relate to these foodat- home purchases. Although the data contain some nutrient data, they are not sufficient to measure how well Americans follow dietary advice or what may motivate them to do so. USDA compiles extensive nutrient and food group databases to support dietary intake studies including the National Health and Nutrition Examination Survey (NHANES). In this study, we use probabilistic and semantic matching techniques to merge the scanner data with the USDA nutrient and food composition databases. As an illustration, we use the new purchase-to-plate “crosswalk”—consisting of matches and conversion factors—to estimate the overall nutritional quality of Americans’ food-at-home purchases. The 2015 Healthy Eating Index (HEI-2015) score for 2013 retail scanner sales is 55 out of 100, indicating that Americans need to substantially improve the healthfulness of their grocery purchases if they wish to follow the Federal Government’s Dietary Guidelines for Americans.