The "mass of tissue at risk" and the myocardial infarct developed was studied in dogs subjected to either 24-h occlusion of the left anterior descending coronary artery or 2-h occlusion followed by 22-h reperfusion. The "mass of tissue at risk" was defined under anaesthesia at the time of occlusion using the microsphere technique. Twenty-four hours later the hearts were removed, sliced transversely and stained with 2,3,5-triphenyltetrazolium chloride to define the infarcted tissue. All myocardial tissue was mapped and cut into small pieces for weighing and radioactive counting. Radioactivity was present in all tissue, including the infarct. In the centre of the infarct, counts remained low and then increased very rapidly with distance just beyond the edge. Tissue at risk from infarction was taken as that with less than 15% of the peak left ventricular (non-ischaemic) counts. A linear relationship was found between the mass of the left ventricular infarct and the left ventricular "mass of tissue at risk". The effect of 22 hours reperfusion was examined by this method and expressed by a regression equation. There was a significant decrease in slope for the regression line of the reperfusion data, (p less than 0.05, analysis of covariance), indicating less infarcted tissue for each gram of underperfused tissue. None of the drug pretreatments explored had any effect on infarct size in the 24-h occlusion model. With reperfusion, propranolol and flunarizine diminished infarct size compared with reperfusion only (p less than 0.05 for reduced slope, the new slope being not significantly different from zero). The effect of diltiazem was not so marked. Thus infarct size can be reduced with pretreatment, as long as the myocardium is reperfused.