In spite of widespread belief among clinicians that caffeinated drinks are linked with palpitations, tachycardia and dysrhythmia there is paucity of documentary evidence. We investigated the arrhythmogenic activity of caffeine in canine model. The alkaloid was given i.v. to perform 51 experiments in 13 anesthetized dogs in three different doses. The Low dose generated significant Sinus (S) bradycardia (70%; p < 0.006, S. arrhythmia (70%; p < 0.02), S. arrest (50%; p < 0.04), Atrial (A) ectopics (40%; p < 0.016), Wandering of pacemaker (WPM) (50%; p < 0.04), and Ventricular premature contractions (VPC--unifocal 40%; p < 0.05) as compared with control ECGs. Medium dose induced significant S. arrhythmia (62%; p < 0.001), A. ectopics (25%; p < 0.01), A. tachycardia (25%; p < 0.01), WPM (25%; p < 0.01), VPCs--unifocal (50%; p < 0.002), multifocal (25%; p < 0.01), couplets (25%; p < 0.01) and interpolated (25%; p < 0.01). High dose of caffeine revealed significant S. arrhythmia (56%; p < 0.002), A. ectopics (44%; p < 0.005), A. tachycardia (32%; p < 0.01), WPM (32%; p < 0.01) and VPCs--unifocal (64%; p < 0.001), multifocal (32%, p < 0.01), couplets (32%, p < 0.01), interpolated (32%; p < 0.01) and Ventricular tachycardia (VT) (20%; p < 0.01). A. flutter and fibrillation each were observed in two experiments only. In conclusion, these data indicate a dose dependent arrhythmogenecity of caffeine. Small dose, mostly, generated, benign arrhythmias due to vagal stimulation. More severe arrhythmias like VT, multifocal VPC. A. flutter and A. fibrillation were generated with higher dose of caffeine. Mechanism remains uncertain as caffeine has multiple actions. Further studies in human beings with normal and compromised myocardium may elucidate arrhythmogenic effects of caffeine.