31P Nuclear Magnetic Resonance (NMR) spectroscopy was used, in combination with biochemical methods, to describe the persisting alterations in energy metabolism provoked by graded normothermic (37 degrees C) global ischaemia, and reperfusion in the isolated perfused rat heart. Graded global ischaemia was induced by adjusting the coronary flow to 0, 1.2, 2.8, or 6.5% of the spontaneous coronary flow in hearts perfused retrogradely under 100 cmH2O (9.807 kPa) perfusion pressure. The 24 min ischaemia was followed by 30 min reperfusion with spontaneous coronary flow. Other series of hearts were perfused with a glucose-free buffer, they were submitted to identical restrictions of coronary flow but for 9 min only with a reperfusion of 20 min. NMR spectra (3 min) were taken throughout the perfusion-ischaemia-reperfusion sequence and used to follow the time-changes in intracellular pH and in the intramyocardial levels of phosphate compounds. Hearts were freeze-clamped at the end of reperfusion to allow for biochemical measurements to be made. Analysis of the results was mainly focused on the energy state at the end of reperfusion. At the end of ischaemia, the extent of the decrease in intracellular pH and the changes in phosphate compound levels were sharply dependent on the degree of coronary flow restriction. In glucose-free perfused hearts, the intracellular acidosis was less than in the presence of glucose. At the end of reperfusion, three kinds of metabolic alterations could be distinguished: 1) those, such as the extent of rephosphorylation of creatine, which were undiscriminative of the acuteness of the previous ischaemia; 2) those, such as the degree of the persisting depletion in ATP (and in the sum of adenine nucleotides), which were directly correlated to the degree of restriction of ischaemic coronary flow; 3) and those which characterised only the most severe conditions of ischaemia, namely a persisting increase in myocardial inorganic phosphate content, a residual shift, albeit slight, of intracellular pH toward acidic values and a displacement of the adenylate charge below control value. The assumption is made that these latter indices can be used to differenciate between reversible and irreversible metabolic damage. An index, calculated from NMR data and correlating well with the adenylate charge, is proposed.