OBJECTIVE--To validate a simple noninvasive method with serial creatine kinase measurements for diagnosis of early patency of the infarct related artery after thrombolytic treatment with streptokinase. To investigate the relation between early patency of the infarct related artery and prognosis. DESIGN--Patients under 76 years of age and seen within six hours of the start of prolonged chest pain and ST segment elevation were treated with streptokinase (1.5 x 10(6) U) intravenously over 30-60 minutes. Blood for measurement of total creatine kinase activity was taken at baseline and at 1, 2, 3, 4, 8, 12, 16, 20, and 24 hours after starting treatment. The rise in enzyme activity at each time from baseline was expressed as a proportion of the total rise from baseline to peak. PATIENTS--Patients studied were in the following groups: (a) Sixty patients took part in a validation study with angiographic determination of patency of the infarct related coronary artery at 2.6 (0.3) hours (mean (SD)) after starting streptokinase. (b) A further 258 patients did not have early arteriography, but data were added to those from the 60 validation patients to find the relation between enzymatically determined early patency of the infarct related artery and 30 day mortality. (c) A further subset of 232 patients with first infarctions (including patients from groups (a) and (b) had angiocardiography at three weeks after infarction, and data were used to investigate the relation between early patency of the infarct related artery and left ventricular function. MAIN OUTCOME MEASURES--Normalised rate of rise of creatine kinase activity at three hours after starting streptokinase in relation to angiographic patency of the infarct related coronary artery at 2.5 hours; 30 day cardiac mortality; and left ventricular function at three weeks in survivors of first infarction. RESULTS--In the validation study, a rise in three hour creatine kinase activity of > 20% of peak occurred in 34/37 patients with initially patent infarct related coronary arteries (sensitivity 92%), and a rise to < 20% of peak occurred in 21/23 patients with initially occluded arteries (specificity 91%). In the prognostic study, 30 day mortality was 2.1% in the 191 patients with three hour creatine kinase > 20% of peak and 8.7% in the 127 patients with three hour creatine kinase < 20% of peak (p < 0.01). Angiocardiography in three week survivors of anterior infarction (n = 95) showed better left ventricular function when three hour creatine kinase was > or = 20% than when it was < 20% of peak (mean (SEM) end systolic volume 71 (5) v 96 (9) ml, p < 0.02; ejection fraction 56% (2%) v 51% (2%), NS). CONCLUSION--Non-invasive determination of early patency of the infarct related artery by the normalised rate of rise of creatine kinase activity at three hours seems to be reliable, and may be prognostically important and of value for use in clinical trials.