Patients with rest pain or acute peripheral arterial thrombosis are known to have impaired endogenous fibrinolysis, which is associated with an increased risk of early vascular graft thrombosis. This risk is exacerbated by the fibrinolytic shutdown which is known to occur after major surgery. Stanozolol, which has been demonstrated to enhance endogenous fibrinolysis, was therefore used in an attempt to prevent this perioperative fibrinolytic shutdown and so enhance graft patency. Twenty-seven patients were randomized to receive either 50 mg stanozolol or placebo intramuscularly 24 h before operation, followed by a 6 week course of either 5 mg stanozolol or placebo orally, twice daily. On the second day after operation, 10-11 MBq of autologous 111indium-labelled platelets were injected, with scanning over the graft on the 3 following days. Despite using a large depot of stanozolol, significant effects, such as raised plasminogen (P less than 0.001), reduced fibrinogen (P less than 0.001) and reduced euglobulin lysis time (P less than 0.001), were not seen until the seventh day after operation, with maximum benefit at 6 weeks. This was reflected in the 111indium-labelled platelet deposition studies. The placebo group had a progressive increase in platelet deposition on all 3 days. In contrast, those receiving stanozolol showed a lower, static picture of deposition. However, these changes did not attain statistical significance. Three patients experienced early graft thrombosis, two in the placebo group and one in the stanozolol group. Only an incomplete inhibition of the perioperative fibrinolytic shutdown was achieved. Much longer preoperative courses are thus required to allow the maximum effect to be present at the most crucial time. At present, perioperative fibrinolytic enhancement does not appear to be a practical proposition, and we must await the development of new safer and more potent agents.