NMR-based metabolomics has gained important insight into the associations between the metabolic status and health, as metabolomics signatures are found in blood, urine, stools, or saliva, differentiating healthy subjects from those affected by diseases or disorders. Although health status has been linked to diet, a measurable fingerprint is rarely found within the metabolome, demonstrating that the diet is curing or, at least, is modifying the subject metabolome away from or closer to a healthy status. The success in finding the correlation between the metabolome and a diet-related disease has, as the main obstacle, the inability to characterize the actual diet followed by the subject. Thus, a big scientific effort has been launched to find metabolite patterns which are characterizing precisely the personal food consumption in order to classify people according to their actual diet. Most of the studies based on NMR-metabolomics are focused on finding biomarkers within the dietary exposome, e.g., originating from food or gut microbiota, without a specific focus on the endogenous metabolome. The main drawback in such approach is a combination of: (i) the actual composition of the meal, (ii) the bioaccessibility of bioactive compounds, and (iii) the processing capability of the gut microbiota. In this chapter, these three aspects are illustrated, where NMR spectroscopy (effectively or potentially) gains relevant information in the discovery of biomarkers for the true food consumption, as a preliminary step in successful "dietary effect studies."