The popular impression that over half of our food does not come from a retail food (grocery) store is based on food expenditure data and is misleading. This research set out to learn where people obtain the food they report eating and to determine whether there are significant differences between people who buy most of their food from retail food stores and those who do not. Research on food consumption often focuses on household expenditures at retail food stores and various types of restaurants, but tracking the volume of various types of foods purchased from various retail places is not well established. The Continuing Survey of Food Intake of Individuals survey for 1994 showed that 72 percent of the volume of food consumed was from retail food stores. Age had the largest impact on where people shopped, and when and how many meals they ate. Income and household composition had relatively little impact. Cluster analysis grouped consumers based on where they obtained their food. The largest cluster, nearly half of the individuals, were labeled the Home Cookers. They obtained 93 percent of their food from stores and account for 59 percent of food sold from retail food stores. The High Service cluster is only 10 percent of the sample, but they consumed 50 percent of the food sold in restaurants and only 6 percent of food sold by grocers. Looking at the diets of people in the various clusters reveal that those in the Fast Food clusters ate less fat than the average of the sample while High Service (restaurant) users ate more fat. Home Cookers ate less than the average amount of meat, eggs, and vegetables.