Temperature modulates the electrophysiological properties of cardiac tissue. The cooling or warming of the myocardium in animal models has been used for various purposes related to research in cardiac arrhythmias. So far, limitations in the methodology did not allow the simultaneous regional modification of temperature in the epicardium together with electrical mapping. This thesis aims to advance the field of the evaluation of the electrophysiological changes induced by temperature in models of isolated and perfused rabbit heart. We have developed a device that includes a high resolution multi-polar electrode capable of modifying epicardial temperature in the recording area. Later, we have determined their most relevant functional and technical features. We have applied an experimental protocol aimed to study the modulating effect of temperature on several electrophysiological parameters. Were studied: the QT interval in sinus rhythm, the onset of repetitive responses (spontaneous depolarizations) and ventricular fibrillation (VF) induction under programmed stimulation, and during VF: the conduction velocity, the wavelength of ventricular refractoriness, the complexity of epicardial activation maps and termination of VF by selective and gradual epicardial cooling. The results show the feasibility of the use of this device for this type of research. This has enabled the induction of heterogeneities in the electrical activation of the tissue: in sinus rhythm the QT interval was prolonged by local hypothermia and reduced by hyperthermia. During VF, cooling reduced conduction velocity whereas warming the tissue increased conduction velocity. Temperature changes modulated the VF wavelength, the fibrillatory interval (mean fibrillatory cycle length) and refractoriness-estimated as the 5th percentile of the fibrillatory interval. There were changes in activation patterns due to temperature, increasing the complexity of the activation maps in hypothermia. Under conditions of programmed stimulation, no association was found between the presence of repetitive responses and the induced heterogeneities, which did not facilitate the induction of VF. Finally, we found that epicardial application of selective and progressive cooling in the perfused heart model used, was not sufficient to terminate VF.