Given the dramatic rise in obesity and related disorders, it is imperative to improve the accuracy of dietary self-monitoring, a cornerstone of treatment. An ambulatory self-monitoring study and a laboratory food estimation study were used to examine (1) the role of eating style (gorging) and weight status (obesity) in dietary underreporting and (2) portion size as mechanism underlying dietary underreporting. Gorging was defined as two or fewer meals per day with at least seven hours between waking and the first meal. Obese was defined as a BMI between 25 and 34.9 kg/m 2. Seventy-six women, ages 19-50 participated. A 2 x 2 (weight by eating style) between groups one-week ambulatory study design was used to examine the accuracy of dietary self-monitoring. Reported energy intake (EI), from a self-monitoring eating diary, was compared to measured energy needs assessed by an ambulatory activity monitor. Accuracy was determined by the Goldberg equation and the ratio of energy intake to energy expenditure (EI:EE). Overweight and gorgers were expected to underestimate EI as compared to normal weight and non-gorgers, respectively. Overweight gorgers were expected to underestimate EI compared to all other groups. i v The effect of portion size on meal size estimation was examined using a laboratory based 2 x 2 x 2 x 2 (weight by eating style by meal size by time) mixed design. Both a regular and large meal size were presented during 2 laboratory visits, scheduled one week apart. Overweight gorgers were expected to underestimate large meals to a greater extent than all other groups. Groups were expected to underestimate large compared to regular meals. Overall, energy needs were greater for obese and gorgers compared to normal weight and non-gorgers. However, there were no differences in report EI among overweight compared to normal weight women, and gorgers reported less EI than nongorgers. This lack of difference in reported EI, a possible indication of dietary underreporting by overweight women, was not confirmed using the Goldberg equation or EI:EE. The Goldberg equation categorized 93.4% of all participants as underreporters, with no differences between groups. Comparing the EI:EE ratio between groups indicated that gorgers underreported compared to non-gorgers. For meal estimation, regular meals were less accurately estimated than large meals. Unexpectedly, all groups overestimated both small and large meals. Few studies to date have examined factors explaining the association between weight and accuracy in reported intake. The role of large portion sizes as a mechanism underlying dietary underreporting was not supported. Future research should continue to focus on understanding mechanisms associated with accuracy of dietary self-monitoring. Knowing why people underreport can lead to improvements in accuracy, increasing our understanding of the relationship between eating behaviors and health, and also improve efforts for weight loss and weight maintenance. v