31P NMR at 72.9 MHz using a 25-mm-diameter phosphorus probe has been used to study correlations among cardiac metabolism, tissue pH, and contractile performance. In all studies, isovolumic left ventricular pressure (LVP) was measured in the paced hearts. Under these conditions, excellent spectra were collected in 5 min, on 6-g rabbit hearts. Good spectra were obtained in 30-sec intervals. Rapid sequential spectra illustrate the metabolic and pH events associated with the onset and recovery of total, global ischemia. Tissue pH was stable for the first minute, but then fell from 7.4 to 6.9 by 6 min, and progressed to a value of 6.4 after 40 min. During the first minute, LVP fell 80%. Metabolites and tissue pH recovered within 6 min of reperfusion, a time when ventricular pressure remained depressed by 50%. These results suggest that both the early fall and initial postischernic recovery of ventricular pressure may not be exclusively regulated by tissue pH, as estimated by NMR. We also used NMR to investigate the metabolic changes associated with regional ischemia. The 31P NMR spectrum of a regionally ischemic, perfused rabbit heart showed two inorganic phosphate (Pi) peaks. Before ligation there was only a single Pi signal at a position corresponding to one of the peaks noted during regional ischemia. Since the resonance frequency of the Pi peak is determined by pH and since ischemia causes acidosis the two signals in the regionally ischemic heart result from Pi at different intracellular pH values in the normal (pH 7.4) and ischemic zones (pH 6.4). And finally, we compared the status of KCl-arrested, ischemic rabbit hearts and non-KCl-treated hearts and correlated the 31P NMR spectra with the ability of the heart to return to normal function following a period of ischemia. A rabbit heart was arrested by perfusing it with 30 mM KCl and was then made globally ischemic; a second heart was made ischemic without KCl arrest. After 40 min of global ischemia the KCl-arrested heart showed a near-normal level of ATP, low Pi and a pH of 7.0. The nonarrested heart, on the other hand, showed low ATP, high Pi, and a pH of 6.4. On reperfusion, the KCl-arrested heart recovered 100% of control function within 5 min but the control heart recovered only 70% of control function after 30 min. The 31P NMR confirms that KCl arrest preserves ischemic myocardial metabolites and suggests that it can be used to test currently untried treatments for functional protection.