N.m.r. is finding increasing application as a non-invasive method of studying metabolism in vivo. 31P n.m.r. is particularly useful for studies of energy metabolism, while other nuclei, in particular 1H, 13C and 19F, can provide additional information about certain aspects of cellular metabolism that are not amenable to investigation by 31P n.m.r. The major problem with n.m.r. is lack of sensitivity. The main implications of this are first that signals can only be detected from metabolites that are present at fairly high concentrations, and secondly that the spatial resolution for localized studies, particularly of internal regions of the body, is limited. N.m.r. could considerably enhance our understanding of metabolic abnormalities in disease, but the precise role of such metabolic studies in a clinical setting remains to be evaluated.