Plasminogen activator inhibitor-1 (PAI-1) plays a central role in the modulation of intravascular thrombosis and thrombolysis. The level of transcription and the plasma level of PAI-1 are in part determined by the 4G/5G polymorphism in the promoter region of the gene. In this study we investigate the effect of 4G/5G polymorphism on the efficacy of reperfusion therapy in acute myocardial infarction (AMI). Two hundred and ninety-three patients were enrolled in two randomized trials comparing stenting plus abciximab with thrombolysis (alteplase alone or alteplase plus abciximab) in AMI. Patients were genotyped for the PAI-1 4G/5G polymorphism with a TaqMan assay. Technetium-99m sestamibi was injected before and 1-2 weeks after reperfusion treatment. Scintigrams were used to calculate the initial perfusion defect, the final infarct size and the salvage index representing the proportion of the initial defect salvaged by reperfusion. An 18-month clinical follow-up was carried out after reperfusion treatment. The distribution of genotypes was as follows: 4G4G in 28.0%, 4G5G in 49.5% and 5G5G in 22.5% of the patients. No significant differences between the three genotypes were detected for the final infarct size (%) of the left ventricle [median (interquartile range); 13.5 (5.0--27.0) for 4G4G patients, 12.0 (5.2--24.6) for 4G5G patients and 16 (7.1--31.2) for 5G5G patients; P=0.36], the salvage index [0.49 (0.25--0.75) in 4G4G patients, 0.47 (0.18--0.73) in 4G5G patients and 0.46 (0.22--0.62) in 5G5G patients; P=0.58] and the mortality 18 months after treatment (8.5% for 4G4G patients, 7.6% for 4G5G patients and 6.1% for 5G5G patients; P=0.85). There was no association in any of the two treatment groups (stenting and thrombolysis) between the 4G/5G genotype and myocardial salvage. The PAI-1 4G/5G polymorphism has no impact on the amount of myocardial salvage achieved by reperfusion with stenting or thrombolysis in patients with AMI.