Abstract Methodological issues that arise when working with scanner data vary depending on whether researchers are using data reported at the store level or household level or are using product-level data aggregated to geographic areas, marketing channels, or demographic categories. Complications arise particularly with underreporting at the store or household level, unreported types of products, and types of variables provided in analysis datasets. To account for common limitations or to prepare data in an appropriate form to address a research question, researchers may need to create their own aggregated data, adjust quantities or prices to make the data more representative or complete, construct appropriate time-series price indices, and determine whether to weight the data.