The measurement of tissue-specific metabolic fingerprints can be of particular interest when investigating disease processes, mechanisms of toxicity, or when knowledge of the metabolic interactions between different organs is required. This chapter presents several optimized protocols for the extraction of metabolites from animal tissues, their analysis by 1H nuclear magnetic resonance (NMR) spectroscopy, and the subsequent spectral preprocessing required for an NMR-based metabolomics experiment. First, the three critical steps in the preparation of tissue extracts for NMR analysis are described, including both a perchloric acid protocol for the extraction of polar metabolites, and a methanol:chloroform protocol for extraction of polar and lipophilic metabolites. Then a series of NMR experiments are described including a standard one-dimensional (1D) 1H NMR study, a 1D 1H Carr-Purcell-Meiboom-Gill spin-echo experiment, and a two-dimensional 1H-1H J-resolved NMR experiment. The advantages and limitations of each experiment for metabolomics research are discussed. Analysis of the resulting NMR datasets is typically conducted in two phases comprising "low level" spectral preprocessing and "high level" multivariate analysis. NMR spectral preprocessing is a critical step that converts raw NMR spectra into an appropriate data format for multivariate analysis. A detailed protocol for preprocessing NMR data, using ProMetab software, is presented. Because a plethora of algorithms exist for multivariate analyses, which can be used to construct classification models or for biomarker discovery, this is beyond the scope of the current chapter.