Diabetes mellitus (DM) promotes cardiac arrhythmias, whereas melatonin confers antiarrhythmic effects. Cellular targets modified by DM and melatonin largely overlap, which warrants testing melatonin as an antiarrhythmic agent in DM. Herein, we evaluated electrophysiological and antiarrhythmic properties of melatonin in the experimental DM model. Experiments were performed in 41 control and 71 diabetic rats. 34 diabetic animals were treated with melatonin (orally, 10 mg/kg daily) for 1 month, and 37 diabetic animals received a placebo. Electrophysiological studies included electrocardiography, in vivo epicardial mapping with induction of ischemia-reperfusion, and patch-clamp studies in ventricular cardiomyocytes. Collagen deposition was assessed in postmortem histological studies. RT-PCR and Western blotting methods were used to estimate Gja1 gene expression and Cx43 protein level, respectively. Melatonin treatment in diabetic animals did not influence the level of glycaemia, but mitigated the progression of myocardial fibrosis. Melatonin treatment prevented the DM-related decrease in conduction velocity, increase in duration and dispersion of repolarization, as well as corresponding changes in their electrocardiographic correlates (QRS, QT, and Tpeak-Tend intervals, respectively). Melatonin also prevented the development of reperfusion ventricular tachycardia/fibrillation. The patch-clamp studies showed that melatonin treatment significantly reduced calcium current, prevented DM-related action potential duration prolongation, but did not modify DM-related increase in sodium current. RT-PCR studies demonstrated that melatonin treatment prevented the DM-related upregulation of Gja1 gene expression and increase in Cx43 level. Melatonin treatment of diabetic rats conferred antiarrhythmic effects associated with electrophysiological changes reflected in ECG parameters, which warrants further testing of melatonin in a clinical context.