Magnetization transfer nuclear magnetic resonance (NMR) provides measurement of the velocity of the creatine kinase reaction in the intact heart. Standard one-pulse NMR spectroscopy coupled with conventional biochemical analyses provides information about the average cytosolic concentrations of ATP, creatine phosphate (CrP), creatine (Cr) and H+ in the heart. By combining these techniques, we tested the hypothesis that the velocity of the creatine kinase reaction in vivo was regulated by changes in cytosolic concentrations of its substrates. We found that the reaction velocity cannot always be predicted from its metabolite levels. We interpreted these observations as support for the hypothesis that flux through the creatine kinase reaction is regulated by metabolite levels in microcompartments formed by localization of creatine kinase isozymes.