To determine the clinical consequences of reocclusion of an infarct-related artery after reperfusion therapy, we evaluated 810 patients with acute myocardial infarction. Patients were admitted into four sequential studies with similar entry criteria in which patency of the infarct-related artery was assessed by coronary arteriography 90 minutes after onset of thrombolytic therapy. Successful reperfusion was established acutely in 733 patients. Thrombolytic therapy included tissue-type plasminogen activator (t-PA) in 517, urokinase in 87, and a combination of t-PA and urokinase in 129 patients. All patients received aspirin, intravenous heparin and nitroglycerin, and diltiazem during the recovery phase. A repeat coronary arteriogram was performed in 88% of patients at a median of 7 days after the onset of symptoms. Reocclusion of the infarct-related artery occurred in 91 patients (12.4%), and 58% of these were symptomatic. Angiographic characteristics at 90 minutes after thrombolytic therapy that were associated with reocclusion compared with sustained coronary artery patency were right coronary infarct-related artery (65% versus 44%, respectively) and Thrombolysis in Myocardial Infarction (TIMI) flow 0 or 1 (21% versus 10%, respectively) before further intervention. Median (interquartile value) degree of stenosis in the infarct-related artery at 90 minutes was similar between groups: 99% for reoccluded (value, 90/100%) compared with 95% for patent (value, 80/99%). Patients with reocclusion had similar left ventricular ejection fractions compared with patients with sustained patency at follow-up. However, patients with reocclusion at follow-up had worse infarct-zone function at -2.7 (value, -3.2/-1.8) versus -2.4 (SD/chord) (value, -3.1/-1.3) (p = 0.016). The recovery of both global and infarct-zone function was impaired by reocclusion of the infarct-related artery compared with maintained patency; median delta ejection fraction was -2 compared with 1 (p = 0.006) and median delta infarct-zone wall motion was -0.10 compared with 0.34 SD/chord (p = 0.011), respectively. In addition, patients with reocclusion had more complicated hospital courses and higher in-hospital mortality rates (11.0% versus 4.5%, respectively; p = 0.01). We conclude that reocclusion of the infarct-related artery after successful reperfusion is associated with substantial morbidity and mortality rates. Reocclusion is also detrimental to the functional recovery of both global and infarct-zone regional left ventricular function. Thus, new strategies in the postinfarction period need to be developed to prevent reocclusion of the infarct-related artery.