OBJECTIVE To determine the hemodynamic effect of vasopressin on coronary perfusion pressure (CPP) in prolonged human cardiac arrest. METHODS A prospective, open-label clinical trial of vasopressin during cardiac resuscitation was performed. Ten patients presenting in cardiac arrest initially received resuscitative measures by emergency physicians according to Advanced Cardiac Life Support (ACLS) guidelines. A central venous catheter for fluid and drug administration and a femoral artery catheter for measurement of CPP (aortic minus right atrial relaxation phase pressures) were placed. When each patient was deemed nonsalvageable, 1.0 mg epinephrine was given and CPP was measured for 5 minutes, followed by a dose of vasopressin (1.0 U/kg). CPP measurements were continued for another 5 minutes. RESULTS The mean duration of cardiac arrest (out-of-hospital interval plus duration of ED ACLS) was 39.6 +/- 16.5 min. There was no improvement in CPP after 1.0 mg of epinephrine. Vasopressin administration resulted in a significant increase of CPP in 4 of the 10 patients. Patients responding to vasopressin had a mean increase in CPP of 28.2 +/- 16.4 mm Hg (range: 10-51.5), with these peak increases occurring at 15 seconds to 4 minutes after administration. The increases in the vasopressin levels after administration did not differ between the responders and nonresponders. CONCLUSIONS In this human model of prolonged cardiac arrest, 40% of the patients receiving vasopressin had a significant increase in CPP. This pilot study suggests that investigation of earlier use of vasopressin as a therapeutic alternative in the treatment of cardiac arrest is warranted.