The quality of survey data obtained from elderly adults has important implications for nutrition status assessment, policy decisions and program development. Self-reported information by adults of all ages is subject to inaccuracies. Accuracy of data obtained from older respondents may be lower than that of younger respondents. Errors made by older adults are largely due to misinterpretation of the questions asked, motivations related to participation and reporting of information, and changes in memory associated with aging and other factors which may impact respondent accuracy. Cooperative efforts by cognitive psychologists, survey researchers and nutritionists are necessary for improving both questionnaire design and the accuracy of information obtained.