The aim was to evaluate the protective effect of verapamil during myocardial ischaemia and reperfusion. In vivo phosphorus-31 (31P) magnetic resonance spectroscopy was performed on rats pretreated with verapamil (mg.kg-1 intraperitoneal) and controls during a 45 min left coronary artery occlusion and 60 min reperfusion. In separate groups of animals, haemodynamic measurements were taken at baseline, during ischaemia, and during reperfusion. Infarct size was determined by staining with triphenyltetrazolium chloride. Female Sprague-Dawley rats were used (control group n = 25, experimental group n = 24). Infarct size was significantly reduced in the verapamil group compared to controls: 9.9(SEM 2.3)%, n = 19 v 28.5(2.7)%, n = 19, p less than 0.001 (infarct % of left ventricular mass). Myocardial phosphocreatine and ATP levels were reduced to similar levels in both verapamil and control animals after 45 min ischaemia: 56.8(3.4)%, n = 10, v 61.4(1.8)%, n = 11 NS; 67.7(2.7)%, n = 10 v 69.7(2.9)%, n = 11, NS (% of baseline value). After 60 min reflow, there was significant recovery of phosphocreatine [91.1(4.2)% of baseline, p less than 0.05] and ATP [86.8(2.7)% of baseline, p less than 0.05] in the verapamil group, but no recovery of high energy phosphates in controls [66.3(2.8), NS; 69.6(2.7), NS]. The left ventricular systolic pressure, heart rate, rate-pressure product, and maximum rate of left ventricular pressure development were similar prior to ischaemia, and during ischaemia in both groups. There was an inverse correlation between infarct size and the degree of phosphocreatine recovery after 60 min of reperfusion (PCr recovery (%) = -0.99 x infarct size (%) + 101; r = 0.91; p less than 0.01; n = 14). Furthermore, in a separate group of animals (n = 9), there was a significant correlation between the size of the ischaemic area at risk and the degree of phosphocreatine decline after 15 min of coronary occlusion (PCr reduction (%) = 0.91 x risk area (%) + 5.6; r = 0.97; p less than 0.01). Pretreatment with verapamil extends the ischaemic time after which reperfusion results in myocardial salvage in this model of ischaemia and reperfusion. This protective effect is independent of the haemodynamic determinants of myocardial oxygen demand and the degree of ATP and phosphocreatine depletion during the ischaemic period. In this model of reversible ischaemia, 31P magnetic resonance spectroscopy is useful for quantitating both the size of the ischaemic region during coronary artery occlusion and infarct size after reperfusion.