The aim of this study was to investigate the ability of troponin I (cTnI) levels to predict myocardial infarction size in patients with ST-segment elevation acute myocardial infarction (STEMI) undergoing primary percutaneous coronary intervention (PCI). In 87 patients with STEMI undergoing primary PCI, serial plasma concentrations of cTnI and alpha-hydroxybutyrate deshydrogenase (HBDH) were measured before PCI and over the following 72 h. Enzymatic infarct size was estimated by the cumulative release of HBDH during the 72 h following PCI (QHBDH72). Delayed radionuclide left ventricular ejection fraction (LVEF) was measured in 63 patients. While cTnI concentrations at admission did not correlate with QHBDH72 or with LVEF, from the 3rd to the 72nd h following PCI, they did correlated with QHBDH72 (P<0.001; R: 0.76-0.86) and with LVEF (P<0.001; R: -0.42 to -0.50). Receiver-operator characteristic (ROC) curve analysis showed that admission concentrations of cTnI could not predict either a large infarct size (i.e., QHBDH72>10 g-eq l(-1)) or a low LVEF (i.e., LVEF<40%). However, 6 h and up until 72 h after PTCA, cTnI concentrations were predictive of large enzymatic infarct size (sensitivity: 91 and 95%, specificity: 90 and 87%, respectively) and of LVEF under 40% (sensitivity: 75 and 77%, specificity: 90 and 78%, respectively). Thus, our study suggests that in contrast with admission cTnI concentration, cTnI levels following primary PCI represent a reliable tool for predicting large enzymatic infarct size and may help in selecting patients with a high risk of low LVEF at 1 month.