The origin of the nuclear magnetic resonance (NMR)-measurable ATP ⇄ Pi exchange and whether it can be used to determine net oxidative ATP synthesis rates in the intact myocardium were examined by detailed measurements of ATP⇄ Pi exchange rates in both directions as a function of the myocardial oxygen consumption rate (MV02) in (1) glucose-perfused, isovolumic rat hearts with normal glycolytic activity and (2) pyruvate-perfused hearts where glycolytic activity was reduced or eliminated either by depletion of their endogenous glycogen or by use of the inhibitor iodoacetate. In glucose-perfused hearts, the Pi → ATP rate measured by the conventional two-site saturation transfer (CST) technique remained constant while MV02 was increased approximately 2-fold. When the glycolytic activity was reduced, the Pi → ATP rate decreased significantly, demonstrating the existence of a significant glycolytic contribution. Upon elimination of the glycolytic component, the measured Pi→ ATP rates displayed a linear dependence on MVO (micromoles of O consumption rate) with a slope of 2.36 ± 0.15 (N = 8, standard error of the mean). This linear relationship is expected if the rate determined by CST is the net rate of ATP synthesis by the oxidative phosphorylation process, in which case the slope must equal the P:0 ratio. The ATP → Pirates and rate:MVO ratios measured by the multiple-site saturation transfer method at two MV02 levels were equal to the corresponding Pi → ATP rates and rate:MVO ratios obtained in the absence of a glycolytic contribution. The following conclusions are drawn from these studies: (1) unless the glycolytic contribution to the ATP ⇄ Pi exchange is inhibited or is specifically shown not to exist, the myocardial Pi ATP exchange due to oxidative phosphorylation cannot be studied by NMR; (2) at moderate MV02 levels, the reaction catalyzed by the two glycolytic enzymes glyceraldehyde-3-phosphate dehydrogenase and 3-phosphoglycerate kinase is near equilibrium; (3) the ATP synthesis by the mitochondrial H+-ATPase occurs unidirectionally (i.e., the reaction is far out of equilibrium); (4) the “operative” P:o ratio in the intact myocardium under our conditions is significantly less than the canonically accepted value of 3.