13C-nuclear magnetic resonance (NMR) spectroscopy provides a new approach to the analysis of metabolic pathways, because it detects an interaction between adjacent 13C nuclei. Previous models of isotope distribution in the tricarboxylic acid cycle were designed for analysis of radioisotope data and did not consider the information provided by 13C-13C coupling. A mathematical model of the tricarboxylic acid cycle was developed that preserves all isotope isomer (isotopomer) information and yields simple relationships between 13C-NMR spectra of glutamate and metabolic parameters under steady-state conditions. With the use of relative peak areas measured from the spectra of tissues supplied with 13C-enriched substrate(s), the relative fluxes through both oxidative (acetyl-CoA utilization) and nonoxidative (anaplerotic) pathways of the tricarboxylic acid cycle can be determined. Furthermore, with the judicious selection of 13C-labeling patterns in a mixture of substrates, direct substrate competition experiments can be performed. The perchloric acid extracts of Langendorff and working rat hearts oxidizing 13C-enriched fatty acids or carbohydrates are analyzed to illustrate this approach, and the importance of measuring the fractional enrichment of the available substrate is demonstrated. The technique can of course be used with all tissues, not just heart, and is applicable to the analysis of in vivo 13C-NMR spectra.