Hypothesis: Alternating intrathoracic pressure by means of a chest cuirass can cause perfusion and ventilation equal to or better than standard cardiopulmonary resuscitation (CPR) for humans in cardiac arrest. Design: Nonrandomized, nonblinded, crossover pilot study. Setting: Large urban emergency department. Subjects: Five adult normothermic, nontraumatic, out-of-hospital cardiac arrest patients unresponsive to standard advanced cardiac life support. Method: Right atrial and aortic catheters were inserted for pressure measurement and blood gas analysis while the patient was receiving standard CPR by a pneumatic compression device (Thumper). The Thumper was then replaced by a chest cuirass (Hayek Oscillator). Pressure and blood gas measurements were then repeated. Results: The coronary perfusion pressure increased from-1.2 ± 8.6 mmHg to 6.2 ± 6.9 mmHg for a mean change of 7.4 ± 3.1 mmHg (p = 0.006). The compression phase gradient increased 10.0 ± 21.9 mmHg (p = 0.364). The venous to arterial Pco 2 gradient decreased 44.5 ± 32.3 mmHg (p = 0.070). The oxygen extraction ratio increased 1.6 ± 9.4% (p = 0.761). The mean arterial Po 2 and Pco 2 changed from 252 to 240 mmHg (p = 0.836) and from 53 to 66 (p = 0.172) mmHg, respectively. Conclusion: The Hayek Oscillator chest cuirass produced a significant improvement in the coronary perfusion pressure. There was a trend for improved systemic perfusion as indicated by an improved compression phase gradient and venous to arterial Pco 2 gradient, although this was not supported by the lack of improvement in the oxygen extraction ratio. The cuirass also adequately oxygenates and ventilates unassisted by positive pressure ventilation.