Objectives: To evaluate the efficacy of conventional threshold levels of coronary perfusion pressure and end-tidal CO, as predictors of resuscitability after prolonged cardiac arrest, Design: Prospective, randomized, controlled animal study, Setting: University research laboratory, Subjects: Twenty-one Sprague-Dawley rats, including three groups of seven animals in each group, Interventions: Ventricular fibrillation was untreated for 9, 12, or 15 mins, After an additional 5-min interval of precordial compression, external direct current defibrillation was attempted, Measurements and Main Results: All animals were successfully resuscitated after 9 mins of ventricular fibrillation but less than one half of the animals were successfully resuscitated after 15 mins of ventricular fibrillation, Each of seven animals survived for 24 hrs after 9 mins of untreated ventricular fibrillation but none of the animals survived after 15 mins of ventricular fibrillation, In this experimental setting, neither coronary perfusion pressure nor end-tidal CO2 produced by precordial compression was predictive of outcomes when the animals underwent progressively longer intervals of untreated cardiac arrest, Conclusions: The efficacy of precordial compression-as measured by coronary perfusion pressure and end-tidal CO2 concentration after prolongation of untreated cardiac arrest-was not overtly compromised, However, the previously established critical threshold levels of coronary perfusion pressure and end-tidal CO2 failed as predictors of resuscitability after prolonged intervals of untreated cardiac arrest.