Acute myocardial infarctions were produced in 11 dogs by ligation of the left anterior descending coronary artery. Twenty-four hours after ligation, 0.35 millimoles per kilogram of Gd-DTPA was injected intravenously, followed by cardiectomy either 90 seconds (3 dogs) or 5 minutes (5 dogs) later. The remaining 3 dogs had cardiectomy without injection of Gd-DTPA at 24 hours after coronary occlusion. The 3 dogs that did not receive Gd-DTPA had longer T1 and T2 relaxation times in infarcted myocardium than in normal myocardium, as measured by a 10.7-MHz magnetic resonance (MR) spectrometer. The T1 and T2 relaxation times of normal myocardium at 90 seconds postinjection of Gd-DTPA were significantly shorter (p less than 0.05) than those of the normal myocardium of animals that did not receive Gd-DTPA. At five minutes postinjection, significantly (p less than 0.01) greater T1 shortening was exhibited in the infarcted myocardium compared with adjacent normal myocardium in the dogs injected with Gd-DTPA. Thus, Gd-DTPA has differential and time-varying effects on relaxation times of normal and infarcted myocardium.