As part of the National Heart, Lung, and Blood Institute multicenter Thrombolysis in Myocardial Infarction Trial, the time to peak plasma creatine kinase (CK) activity as a marker of reperfusion in 272 patients with validated acute myocardial infarction was analyzed. Patients were treated with either tissue-type plasminogen activator or streptokinase by intravenous administration. All patients underwent acute coronary angiography. The infarct-related artery was identified and thrombolytic therapy administered. Reperfusion at 90 minutes was documented by angiography. CK was determined before institution of therapy and every 4 hours thereafter for the first 24 hours. Patients were classified into 3 groups for comparative purposes: group 1--occlusion with no reperfusion (n = 119); group 2--occlusion with reperfusion (n = 98); and group 3--subtotal occlusion (n = 55). Early (within 4 hours after treatment) and late (more than 16 hours after treatment) peaking of CK differentiated patients with drug-induced perfusion from those without reperfusion. Although peak CK between 5 and 11 hours after drug treatment did suggest perfusion through the infarct-related artery, it did not differentiate between drug-induced and spontaneous reperfusion. Clinically, early peak CK is a useful noninvasive means of assessing coronary artery patency. However, in clinical trials assessing drug therapy, the use of peak CK may overestimate drug effectiveness by including patients with spontaneous reperfusion.