Although arginine vasopressin (AVP) has been shown to be a promising drug during cardiopulmonary resuscitation (CPR), concern has been raised about the potential for AVP-mediated vasoconstriction of the coronary arteries. In a prospective, randomized laboratory investigation employing an established porcine model, the effects of AVP on haemodynamic variables, left anterior descending (LAD) coronary artery cross sectional area employing intravascular ultrasound (IVUS), and return of spontaneous circulation were studied. During sinus rhythm, the LAD coronary artery cross sectional area was measured by IVUS at baseline, and 90 s and 5 min after AVP (0.4 U/kg IV). Following a 60 min recovery, ventricular fibrillation was induced. At 4 min, chest compressions were initiated; AVP (0.4 U/kg IV) was injected at 5.5 min, and defibrillation performed at 8 min. LAD coronary artery cross sectional area was measured by IVUS at the pre-arrest baseline, 90 s after drug injection during CPR, and 5 min after return of spontaneous circulation. Compared with baseline, the mid-LAD coronary artery cross sectional area increased significantly (P<.05) 90 s and 5 min after AVP administration (9.2+/-.5mm2 versus 10.7+/-.6mm2 versus 11.7+/-.6mm2, respectively) during normal sinus rhythm. Similarly during ventricular fibrillation and CPR plus AVP, the mid-LAD coronary artery cross sectional area increased at 90 s after AVP compared with baseline (9.5+/-.6mm2 versus 11.0+/-.7mm2; P<.05). Moreover, the cross sectional area increased further 5 min after return of spontaneous circulation (9.5+/-.6mm2 versus 14.0+/-.8mm2, P<.05). In conclusion, in this experimental model with normal coronary arteries, AVP resulted in significantly increased LAD coronary artery cross sectional area during normal sinus rhythm, during ventricular fibrillation with CPR, and after return of spontaneous circulation.