Background: Intravenous unfractionated heparin (UFH) is recommended in ST-elevation acute myocardial infarction (STEMI), following thrombolysis with fibrin-specific agents. Subcutaneous low-molecular-weight heparins (LMWH), previously proven effective in non-ST-elevation acute coronary syndromes, have been recently investigated in the setting of STEMI. We aimed at evaluating the current level of evidence supporting the use of LMWH in STEMI. Methods: A Medline search of the English language literature between January 1995 and December 2005 was performed and randomized clinical trials comparing LMWH to either placebo or UFH in conjunction with thrombolysis were selected. Results: About 26,800 patients treated with various thrombolytic regimens were included in 12 randomized clinical trials. Dalteparin was superior to placebo on left ventricular thrombosis/arterial thromboembolism, with no significant effect on the early patency rate of the infarct-related artery (IRA). Compared to UFH, dalteparin had no significant effect on clinical events and on the IRA late patency, although less thrombus was present. Enoxaparin was superior to placebo on the medium-term death/reinfarction/angina rate and late IRA patency, and superior also to UFH on in-hospital and medium-term occurrence of death/reinfarction/angina. The effect of enoxaparin on IRA patency rate was not univocal. Compared to placebo, reviparin significantly reduced early and medium-term mortality and reinfarction rates, without a substantial increase in overall stroke rate. As regards safety, bleedings were more frequent than placebo and comparable to UFH in LMWH groups, with the exception of the pre-hospital ASSENT-3 PLUS trial, where in elderly patients, enoxaparin had an incidence of intracranial hemorrhage twice higher than UFH. Conclusions: In-hospital subcutaneous administration of dalteparin, enoxaparin or reviparin, as an adjunct to various thrombolytics in STEMI, appears feasible and at least as effective and safe as intravenous UFH. Before LMWH might be recommended, however, some yet unresolved issues (i.e. use in elderly patients, in severe renal insufficiency, in association with glycoprotein IIb/IIIa inhibitors and during interventional procedures), need to be addressed.