Clinical feeding studies offer an excellent opportunity to compare actual and self-reported amounts of foods eaten. We used participants in a multicenter clinical trial, who received well-controlled experimental diets, to compare differences in methods of estimating portions of foods. We examined diet recalls of 18 females and 25 males (aged 22 to 65yrs) participating in the Dietary Effects on Lipoproteins and Thrombogenic Activity (DELTA) study. Three 24-hour recalls were collected by computer-assisted telephone interview on random days using Minnesota Nutrition Data System (NDS). Participants were encouraged to use the 2D Food Portion Visual (FPV) (Nutrition Consulting Enterprises, Framingham, MA) to estimate food portions, but also reported foods in common household measures (CHM), units, dimensions, or weights. The proportions of recalled to actual intakes were similar for the various methods. Foods were categorized according to similarities in the methods of deriving portion estimations. Categories included beverages, side dishes (e.g., cooked vegetables, canned fruits, and rice), meats, cereals, condiments, breads, mixed dishes, and snacks (peanuts, pretzels). Mean gram weights of recalled foods were calculated and presented as percentages of actual gram intakes. Males were almost twice as likely to estimate portion sizes using FPV than females (1294 vs 739) and females used CHM more often than males (603 vs 340). Males were more accurate with estimations using FPV than females when reporting side dishes, meats, mixed dishes, breads, and snacks; females were more accurate using FPV for cereals and condiments. Females were more accurate using CHM than males for meats and cereals; males were more accurate using CHM than females for mixed dishes and snacks. Males and females were similar when using CHM to estimate beverages, condiments, and side dishes. The results indicate that in our sample there are gender differences in the type and accuracy of methods used for portion size estimations. These data have implications for recommendations for portion size aids used in collecting dietary recalls.