Granulocytes, especially neutrophils, are recruited in myocardium during the evolution of acute myocardial infarction. Because the neutrophil reaction is most intense during reperfusion and because these cells are a rich source of toxic oxidant species and proteolytic enzymes, it has become a widely held view that neutrophils are an important mechanism of myocardial injury extension during reperfusion. However, on close examination the evidence underlying this contention is equivocal. The basic experimental situation can be summarised thus. (1) All forms of reperfusion injury (i.e., cytotoxic or lethal cell injury, myocardial stunning, endothelial dysfunction, and reperfusion-induced arrhythmias) can be observed in neutrophil-free conditions. (2) "Anti-neutrophil" interventions (e.g., anti-inflammatory drugs, adenosine, anti-neutrophil antisera, leukocyte filters and inhibitors of the various pathways of neutrophil adhesion) do not consistently prevent reperfusion injury and they certainly do not consistently limit infarct size. (3) The time course of neutrophil accumulation in post-ischaemic myocardium may be different to the time course of injury. (4) Despite more than two decades of research, no double-blind, randomised controlled clinical trial assessing an anti-neutrophil therapy in myocardial infarction has yet reported a positive benefit that is attributable to inhibition of neutrophil recruitment. The evidence weighs against a pivotal role of neutrophils as a causal factor in most forms of ischemia-reperfusion injury. An exception may be microvascular injury and capillary plugging leading to the "no-reflow" phenomenon but even here the evidence suggests that the extent of neutrophil accumulation and microvascular injury is determined by, rather than a cause of, myocyte necrosis.