We use FDA surveys on awareness of diet-disease relationships to estimate a probability model of awareness. We apply the model to respondents of USDA's 1985-88 food consumption surveys to estimate a predicted probability of awareness, an explanatory variable in the multivariate analysis of fat intake. Despite systematic changes in food behavior associated with diet-disease awareness, women with higher awareness probabilities showed no greater reduction in fat intake than others. Difficulties in making effective food substitutions may be due to insufficient knowledge about the relative fat content of different food groups. More research is needed to understand the complex link between diet-disease awareness and dietary practices.