The aim of this study was to investigate the effect of the cytokine, r-metHuG-CSF, in a rat model of ischaemia-reperfusion (IR) injury and the pathophysiological mechanism involved. The administration of r-metHuG-CSF (20 (g/kg, s.c.) 4 h prior to either 100 min or 2 h of tourniquet ischaemia to the upper thigh significantly improved the viability of skeletal muscle after 24 h reperfusion compared with saline-treated rats (P < 0.05). Administration of r-metHuG-CSF earlier (24 h before ischaemia) or later (immediately before ischaemia) had no protective effect. At the dose used, r-metHuG-CSF caused a three-fold increase in the level of circulating blood neutrophils and a modest but significant increase in the neutrophil content of ischaemic muscle after 24 h reperfusion. Reduction of neutrophils to 1.4% of normal levels by cyclophosphamide (150 mg/kg, i.p.) prior to injury had no significant effect on the survival of muscle subjected to 2 h ischaemia and 24 h reperfusion or on the protective effect of r-metHuG-CSF. IR injury to skeletal muscle was accompanied by a time-dependent increase in plasma TNFalpha levels during the first 8 h of reperfusion and the increase was reduced significantly by pretreatment with r-metHuG-CSF. However, a similar time-dependent increase in plasma nitrite/nitrate levels was unaffected by pretreatment with r-metHuG-CSF. These findings suggest that the protective effect of r-metHuG-CSF may be mediated by the attenuated release of TNFalpha and indicate that the level of neutrophils in either blood or injured tissue does not influence significantly the viability of rat muscle after IR injury.