The negative results of glucose-insulin-potassium (GIK) in the very large CREATE-ECLA trial that studied 20,201 patients with ST-elevation acute myocardial infarction (AMI), are disappointing and warrant thorough evaluation. We attempt to put the new data into perspective and uncover the serious flaws in the trial design, otherwise the whole metabolic concept will be disparaged. The crucial issue, developed from basic science data, is that GIK should be initiated very early, before, or at the time of reperfusion. Another problem with CREATE-ECLA is that the mortality in Killip class 1 reperfused patients was 7.1%, much higher than that of a recent Dutch study in which mortality was only 1.2%. Nonetheless, there was a strong trend towards a lower mortality in the sub-groups that received the best reperfusion therapy in CREATE-ECLA, as well as in the first of two rather small Dutch GIK trials. In the future, the ideal protocol to test would be if GIK were given in the ambulance as the patient is being transported to a specialized centre of percutaneous coronary intervention (PCI), with the aim of expanding the time window between pain onset and actual PCI.