OBJECTIVE To assess the effects of graded doses of vasopressin vs. saline on median fibrillation frequency and defibrillation success in a porcine model of cardiopulmonary resuscitation. DESIGN Prospective, randomized, controlled trial. SETTING Animal laboratory in a university medical center. SUBJECTS Twenty-eight domestic pigs (body weight between 26 and 31 kg), aged 12 to 14 wks. INTERVENTIONS AND MAIN RESULTS After 4 mins of ventricular fibrillation and 3 mins of closed-chest cardiopulmonary resuscitation, the animals were allocated to receive either 0.2 U/kg of vasopressin (n = 7), 0.4 U/kg of vasopressin (n = 7), 0.8 U/kg of vasopressin (n = 7), or 10 mL of saline (n = 7, control group). Using radiolabeled microspheres, myocardial blood flow rates during cardiopulmonary resuscitation-before drug administration and 90 secs and 5 mins after drug administration-were as follows in the four groups (mean +/- SEM): 18.8 +/- 0.9, 17.2 +/- 1.1, and 14.6 +/- 1.4 mL/min/100 g in the control group; 17.8 +/- 2.2, 49.6 +/- 6.3 (p < .01 vs. control group), and 29.4 +/- 3.1 mL/min/100 g (p < .05 vs. control group) in the group receiving 0.2 U/kg of vasopressin; 17.1 +/- 1.0, 52.4 +/- 7.5 (p < .01 vs. control group), and 52.2 +/- 5.8 mL/min/100 g (p < .001 vs. control group) in the group receiving 0.4 U/kg of vasopressin; and 18.1 +/- 1.6, 94.9 +/- 9.2 (p < .001 vs. control group), and 57.2 +/- 6.3 mL/min/100 g (p < .001 vs. control group) in the group receiving 0.8 U/kg of vasopressin. Using spectral analysis, median frequencies of ventricular fibrillation-before drug administration and 90 secs and 5 mins after drug administration-were as follows in the four groups: 9.6 +/- 0.4, 8.5 +/- 0.8, and 7.2 +/- 1.0 Hz in the control group; 9.7 +/- 0.5, 12.9 +/- 0.8 (p < .01 vs. control group), and 12.7 +/- 0.8 Hz (p < .001 vs. control group) in the group receiving 0.2 U/kg of vasopressin; 10.3 +/- 0.2, 12.7 +/- 0.9 (p < .01 vs. control group), and 12.8 +/- 0.7 Hz (p < .001 vs. control group) in the group receiving 0.4 U/kg of vasopressin; and 10.0 +/- 0.9, 14.1 +/- 0.9 (p < .001 vs. control group), and 12.5 +/- 0.9 Hz (p < .001 vs. control group) in the group receiving 0.8 U/kg of vasopressin at the same points in time. Median frequency before the first defibrillation attempt was 12.3 +/- 0.4 Hz in the resuscitated animals (n = 19) and 8.2 +/- 1.2 Hz in the nonresuscitated animals (n = 9) (p < .001). CONCLUSIONS This study contributes to the characterization of the effect of increasing global myocardial blood flow on median fibrillation frequency after administration of graded doses of vasopressin in a porcine model of ventricular fibrillation. Interventions such as vasopressor treatment that increase fibrillation frequency improve the chance of successful defibrillation.