Research aim: Investigate the generation of reactive oxygen species in human myocardium after its revascularization by coronary artery bypass graft surgery and probe the possibilites of their reduction. Materials and methods: In order to compair the level of reactive oxygen species produced in the group of patients with developed coronary artery collaterals and the group without them, samples were taken from a total of 11 patients. During the performance of the off-pump CABG, tissue samples of the subepicardial myocardium of the left ventricle were collected by the needle biopsy method on two occasions - before the installation of the bypass (thus in the phase of ischemia) and 5 minutes later after installation of the bypass (in the reperfusion phase). The development of coronary collateralls was assessed using coronary angiography, and evaluated by the Rentrop system. Immediately after collection, the samples were immersed in liquid nitrogen, which allowed the quantification of ROS present in situ at that time. Measurements were then made using the electron paramagnetic resonance (EPR) method. In order to measure the level of ROS via EPR, a CMH indicator (1-hydroxy-3-methoxycarbonyl-2,2,5,5- tetramethylpyrrolidine) was added to the sample, which changes in the reaction with ROS, resulting in stable paramagnetic CM• (3-methoxycarbonyl-proxyl nitroxide) detected by EPR device. In a subgroup of 5 patients in the phase of ischemia (before bypass placement), an additional tissue sample was taken for the analysis of ROS production in vitro. These samples were immediately immersed in the cardioplegic solution and delivered to the laboratory where they were prepared for the experiment in the tissue homogenization device (to which the respiratory solution was added). The homogenate thus prepared was transferred to an experimental chamber (in a device for measuring mitochondrial functions) with the addition of metabolic substrates and the Amplex Red (a detection system that enables the assessment of ROS formation). Amplex Red binds hydrogen peroxide, forming the product resorufin, which has a fluorescence ability that we detect in the system and from which we calculate the level of produced peroxide. Results: Using the EPR method, the level of ROS was measured based on proportionality with paramagnetic CM•. There was no difference in ROS between ischemia and reperfusion when all patients were pooled together. However, in group of patients with no collaterals (Rentrop = 0), reperfusion induced 56% increase in ROS (P < 0.05). In patients with developed collaterals (Rentrop ≥ 2), no change in ROS levels was found. In the metabolic chamber for measuring mitochondrial function, the dynamics of ROS generation during oxidative phosphorylation was first measured, which turned out to be relatively low. 57 The myocardial tissue was then exposed to the metabolic conditions of simulated reperfusion (the main ones being highly reduced ubiquinol and a high proton gradient across the inner mitochondrial membrane) during which ROS production was significantly increased. When the sample was exposed to the same conditions of simulated reperfusion, but in the presence of metformin, ROS production was significantly lower. Conclusions: During reperfusion of ischemic tissue, there is an increase in the production of ROS by the mechanism of reverse electron transfer. The increase in ROS production under conditions of simulated reperfusion is reduced by the administration of metformin. In patients undergoing CABG, there is a significant increase in ROS in the reperfusion phase in the group that does not have developed collateral blood vessels. Therefore, the possibility of using metformin for the purpose of prophylaxis of the development of reperfusion injury opens up, ideally before subjecting patients to revascularization methods, from which those patients who do not have developed coronary collateral circulation will benefit the most.