To assess the importance of contraction band necrosis (CBN) in patients with acute myocardial infarction (AMI) treated with selective intracoronary thrombolysis, CBN, coagulation necrosis, and infarct size (expressed as CBN + coagulation necrosis) were analyzed quantitatively in 16 autopsied hearts. Intracoronary thrombolysis was performed from 2 to 6 hr after the onset of AMI, and the time from the onset of AMI to death was 7 to 168 hr. Cineangiography revealed no evidence of good collateral circulation in any of the patients. The 16 patients were classified into three groups: six patients with successful thrombolysis (100% to 99% stenosis, group I), five patients with unsuccessful thrombolysis (100% to 100%, group II), and five patients with 99% stenosis before thrombolysis (group III). Among the three groups, there were no significant differences in the time from the onset of AMI to thrombolysis, the time from the onset of AMI to death, the cause of death, or the degree of collateral circulation. The percentage of the risk area involved by the infarct in group 1 (82 6%) was similar to that in group 11 (80 + 1 1 %). Infarct size was not reduced in group I because collateral circulation was not good and because the degree of recanalization after thrombolysis was 1%. However, the percentage of the infarct area with CBN was significantly higher in group 1 (20 + 9%) then in group 11 (3 + 3%). This finding shows that diffuse CBN occurred after reperfusion in patients with AMI treated with thrombolysis. In group III, three hearts had a smaller infarct area than that in group I or II and a larger area-of CBN than that in group II. In the other two hearts, the infarct area was similar to that in groups I and II and the CBN area was as small as in group II. We believe that in the former three cases, spontaneous recanalization occurred before thrombolysis and a considerable portion of the risk area was salvaged. In all three groups, the distribution of CBN was mainly in the peripheral zone of the infarct area. Most of the noninfarct tissue areas in the risk area were in the outer third of the left ventricular wall. These findings reflect the transmural gradient of ischemic cellular damage in human AMI. The degree of ischemia immediately before reperfusion is milder in the areas of CBN than in the areas of coagulation necrosis in human AMI. In conclusion, a method should be developed to protect the heart against CBN to reduce infarct size. Circulation 76, No. 5, 981-989, 1987. SINCE Rentrop et al.' first described selective intracoronary thrombolysis in patients with acute myocardial infarction (AMI) in 1979, the technique has been used more and more widely. An