Objective To study the effects of patency of infarct-related coronary artery on acute outcome in cases of acute myocardial infarction undergoing urokinase therapy. Methods One thousand one hundred and thirty-eight cases of acute myocardial infarction admitted in 37 collaborative hospitals were given urokinase intravenously with a standard program. Patency of infarct-related coronary artery assessed by uniform criteria occurred in 757 patients (66.5%) and non-patency in 381 patients (33.5%). Four-week mortality and frequency of various severe complications were compared between patency patients and non-patency patients. Results The 4-week mortality in patients with patency was 3.4% versus 21.8% in those without patency (P < 0.001). Frequencies of various severe complications such as heart failure, shock, ventricular fibrillation, bundle branch blocks and mechanical complications (P < 0.01), and complete atrioventricular block (P < 0.05) were much higher in patients without than in those with patency. In large sized infarcts such as high lateral plus anterior wall infarction the difference in 4 week mortality was even more obvious between patients with and without patency (2.6% vs 32.7%, P < 0.001). Using Cox proportional hazards regression model analysis, patency of infarct-related coronary artery was defined as the most important independent predictor of 4-week mortality. The risk ratio for occluded versus parent infarct-related coronary artery was 6.69. Conclusion On univariate and multivariate analyses, patency of the infarct-related coronary artery was the most important prognostic factor for acute outcome in acute myocardial infarction.