Biomarkers of food consumption are important indicators for clinical and feeding studies that establish our understanding of nutrition. Carbohydrates make up the largest component of a healthy diet. Biomarkers that are derived directly from carbohydrates are ideal. However, there is a fundamental research gap in that carbohydrate structures of specific foods that are commonly consumed are unknown. The long-term goal is to determine specific glycan biomarkers of dietary intake for all commonly consumed foods. The objective of the current application is to develop a detailed carbohydrate structure library of foods in terms of their monosaccharide and linkage compositions and to evaluate carbohydrate structures for their utility as biomarkers of dietary intake. Our central hypothesis is specific foods have unique carbohydrate structures that can serve as biomarkers of dietary intake. This hypothesis has been formulated on the basis of preliminary studies in which (a) detailed carbohydrate structures have been determined for hundreds of foods and (b) fecal glycans, microbes, and metabolites differed by dietary intervention of weaning foods in infants. Enzymes in the gut from human and their complement microbes are specific to structural carbohydrate features. Therefore, the specific aims are to (1) construct a glycomic library with paired analytical platform for food and feces, (2) conduct a microbial gene marker library using bioreactor studies and (3) test the predictive value of glycan and microbial biomarkers in a human feeding study. The approach is innovative, bringing unprecedented high-resolution carbohydrate content analysis—monosaccharide and linkage analysis of foods—with multiple omic measurements to biomarker prediction of dietary intake. The proposed research is significant and impactful because a high-resolution glycan library of foods together with glycan-microbe biomarker products of those foods will transform future clinical trials in which diet is an essential component.