The second Leicester Intravenous Magnesium Intervention Trial (LIMIT-2) examined the effect of an intravenous regimen of magnesium sulphate in 2316 patients with suspected acute myocardial infarction. Treatment, according to a double-blind randomised protocol, was started with a loading injection, before any thrombolytic therapy, and continued with a maintenance infusion for a further 24 h. Cause-specific mortality of randomised patients has now been examined over 1.0-5.5 (mean 2.7) years of follow-up. Mortality rate from ischaemic heart disease was reduced by 21% (95% CI 5-35%, p = 0.01) and all-cause mortality rate reduced by 16% (2-29%, p = 0.03) in magnesium-treated patients. Magnesium protects the contractile function of the myocardium from reperfusion injury ("stunning") in experimental models; this observation accords with the 25% (7-39%, p = 0.009) reduction in early left ventricular failure in the magnesium group of LIMIT-2. For such protection to occur, magnesium must be raised by the time of reperfusion since the injury is immediate. In the clinical context the timing of magnesium treatment in relation to thrombolytic therapy or spontaneous reperfusion is likely to be critical. The early benefits of this simple and safe intervention are reflected in improved long-term survival.