ST-segment resolution(STR) is a surrogate end point in reperfusion trials of acute myocardial infarction, but there are few data regarding the optimum methods of measurement, clinical predictors, and correlation with late cardiac mortality. Consecutive patients(n=1,005) who had acute myocardial infarction and< 2 mm ST-segment elevation controlled with primary percutaneous coronary intervention(PCI) constituted our study group. Follow-up was obtained in 97%of patients at a median of 6.2 years. STR measured as maximum ST-segment elevation after PCI provided better discrimination of late cardiac mortality than did STR measured as percent resolution. Complete STR(< 1.0 mm ST-segment elevation after PCI) was achieved in only 42%of patients. Anterior infarction, Killip’s class 3 to 4, and Thrombolysis In Myocardial Infarction flow grades< 2 before PCI and< 3 after PCI were strong independent predictors of partial or poor STR. STR(complete [< 1.0 mm] vs partial [1.0 to 2.0 mm] vs poor [ >2.0 mm]) correlated with in-hospital mortality(4.0%vs 6.7%vs 11.6%, p=0.005), reinfarction(1.4%vs 3.4%vs 6.1%, p=0.01), and late cardiac mortality(17%vs 25%vs 44%, p< 0.0001). Correlation with late mortality was stronger for nonanterior than for anterior infarction. Poor STR was a strong independent predictor of late mortality(hazard ratio 1.63, 95%confidence interval 1.06 to 2.50, p= 0.028), even after adjusting for Thrombolysis In Myocardial Infarction flow. These data support the use of STR as a simple method to stratify patients by risk after primary PCI for acute myocardial infarction and support the use of STR as a surrogate end point in reperfusion trials of acute myocardial infarction. ST-segment resolution (STR) is a surrogate end point in reperfusion trials of acute myocardial infarction, but there are few data regarding the optimum methods of measurement, clinical predictors, and correlation with late cardiac mortality. Consecutive patients (n = 1,005) who had acute myocardial infarction and <2 mm ST-segment elevation controlled with primary percutaneous coronary intervention (PCI) configured our study group. Follow-up was obtained in 97% of patients at a median of 6.2 years. STR measured as maximum ST-segment Elevation after PCI provided better discrimination of late cardiac mortality than did STR measured as percent resolution. Complete STR (<1.0 mm ST-segment elevation after PCI) was achieved in only 42% of patients. Anterior infarction, Killip’s class 3 to 4, and Thrombolysis In Myocardial Infarction flow grades <2 before PCI and <3 after PCI were strong independent predictors of partial or poor STR. STR (complete [<1.0 mm] vs partial [1.0 to 2.0 mm] vs poor [> 2.0 (vs 4% vs. 6.1%, p = 0.01), and late cardiac mortality (17% vs 25% vs 44%, p <0.0001) Correlation with late mortality was stronger for nonanterior than for anterior infarction. Poor STR was a strong independent predictor of late mortality (hazard ratio 1.63, 95% confidence interval 1.06 to 2.50, p = 0.028), even after adjusting for Thrombolysis In Myocardial Infarction flow. These data support the use of STR as a simple method to stratify patients by risk after primary PCI for acute myocardial infarction and support the use of STR as a surrogate end point in reperfusion trials of acute myocardial infarction infarction.