Time Domain 1 H Nuclear Magnetic Resonance (TD NMR), also known as Low- Field NMR, is an extremely useful technique for measuring mobile water populations and fat protons in food and feed. The bulk constituents such as water, fats and carbohydrates can easily be detected and quantified with virtually no bias. The chemical and physical information gained from TD NMR experiments does require adequate data-analysis techniques in order to establish modeldriven approaches for hypothesis testing, for rapid quantitative applications, as well as for explorative multivariate methods during hypothesis generation. Low field time domain relaxation data is characterized by a lack of structure and selectivity, normally being composite exponential decay functions which are direct functions of the mobility and/or compartmentalization of the molecular systems under study. The signal intensity at a given time is a weighted-proportional to the contributions from the measured substances, i.e. a slow-relaxing component contributes relatively more than a fast-relaxing component. In order to fully exploit in detail the quantitative structural and compositional information produced by TD NMR experiments multivariate data analysis is required. In this chapter different quantitative strategies-namely ratio fitting, discrete exponential fitting, POWERSLICING, multivariate curve resolution, partial least squares regression and inverse Laplace transformation combined with regression - will be demonstrated and compared. This qualitative and quantitative comparison will be based on a data set aimed at predict dry matter contents in potato tubers.