Using consumer panel data we explore the impact of food package size on food-at-home consumption. We exploit food manufacturer package downsizing strategy to track shifts in household purchase volume before and after package size changes. Focusing on shelf-stable tuna and peanut butter markets, we design a difference-in-difference analysis to compare the changes in purchase volume of products that are affected by package downsizing (treatment group) to the changes in purchase volume of products that are not affected by package downsizing (control group). Our main finding is that on average smaller package size significantly reduced household purchase volume in both product categories. This result implies that package size is positively correlated with food-at-home consumption, which is consistent with results of the experimental studies showing that larger package sizes lead to higher usage volume compared with smaller package sizes.