To investigate the dynamic control of cardiac ATP synthesis, we simultaneously determined the time course of mitochondrial oxygen consumption with the time course of changes in high-energy phosphates following steps in cardiac energy demand. Isolated isovolumically contracting rabbit hearts were perfused with Tyrode's solution at 28°C (n=7) or at 37°C (n=7). Coronary arterial and venous oxygen tensions were monitored with fast-responding oxygen electrodes. A cyclic pacing protocol in which we applied 64 step changes between two different heart rates was used. This enabled nuclear magnetic resonance measurement of the phosphate metabolites with a time resolution of =2 seconds. Oxygen consumption changed after heart-rate steps with time constants of 14±1 (mean±SEM) seconds at 28°C and 11±1 seconds at 37°C, which are already corrected for diffusion and vascular transport delays. Doubling of the heart rate resulted in a significant decrease in phosphocreatine (PCr) in inorganic phosphate (Pi) content, although oxygen supply was shown to be nonlimiting. The time constants for the change of both Pi and PCr content, =5 seconds at 28°C and 2.5 seconds at 37°C, are significantly smaller than the respective time constants for oxygen consumption. The changes in phosphate metabolites during changes in oxygen consumption suggest that regulation of oxidative phosphorylation could occur partly via products of ATP hydrolysis, but the unequal time constants of PCr and oxygen consumption suggest that other regulatory mechanisms also play a role. These dissimilar time constants further suggest that there might be an appreciable transient contribution of nonaerobic, presumably glycolytic, ATP synthesis to buffer the high-energy phosphates during fast transitions in cardiac work. (Circ Res. 1994;75: 751-759.) Key Words * myocardial energy metabolism * rabbit hearts * mitochondrial control * 31P nuclear magnetic resonance spectroscopy * oxidative phosphorylation changes in high-energy phosphates and Pi despite increases in myocardial oxygen consumption suggested that the phosphate metabolites may not be the primary regulators of cardiac mitochondrial respiration.The absence of a change in ATP or phosphocreatine (PCr) after a change in metabolic demand would also mean that mitochondrial ATP production and hence oxygen consumption must adapt immediately to a change in cytosolic ATP hydrolysis. In the isolated perfused rabbit heart, this is not the case, because oxygen consumption adapts to a new work load with a mean response time of at least 6 seconds.