1. The effects of the nitric oxide synthase (NOS) inhibitors, NG-nitro-l-arginine-methyl ester (l-NAME), nitroiminoethyl-l-ornithine and S-methylisothiourea on skeletal muscle survival following 2 h of tourniquet ischaemia and 24 h of reperfusion were compared with those of the antiinflammatory steroid, dexamethasone. 2. Administration of each of the NOS inhibitors or dexamethasone 30 min before reperfusion reduced the degree of skeletal muscle necrosis 24 h after reperfusion. 3. The influence of timing of drug administration was investigated. l-NAME administered 30 min before reperfusion, at 3 h after reperfusion, but not thereafter, significantly improved muscle survival compared with saline-treated controls. Dexamethasone administered 30 min before, or at 3 or 8 h after reperfusion, but not at 16 h, significantly improved muscle survival, but neither agent had protective effects when administered before ischaemia. 4. After 8 h of reperfusion of ischaemic skeletal muscle, cell-free homogenates contained Ca2+-independent (inducible) NOS activity which was reduced in dexamethasone-treated (2.5 mg/kg) rats. Furthermore, inducible NOS mRNA levels, as detected by reverse transcriptase-PCR, were increased after 8 h of reperfusion in saline, but not in dexamethasone-treated rats. 5. These data suggest a significant deleterious effect of endogenous NO which may be restricted to the first 3 h of the reperfusion phase of ischaemia-reperfusion injury, and raise the possibility of effective treatment of incipient reperfusion injury, even after several hours of reperfusion.