We present two patients with ventricular fibrillation (VF) during acute myocardial infarction (AMI). First patient had torsade de pointes ventricular tachycardia episodes degenerating into VF 10 days after AMI treated with primary angioplasty. Second patient had multiple episodes of VF during the first day of AMI. He showed ST-segment elevation resembling Brugada pattern. We hypothesise that shape of ST-segment elevation during AMI might be important in assessing risk for VF. We propose that such 'arrhythmogenic' ST elevation might result more from systolic rather than diastolic current of injury during AMI. We discuss genetic predispositions (latent channellopathies) for VF during AMI.