tinuous thermodilution technique) was positioned in the coronary sinus (CSQ). VF was induced electrically and countershock (400 J) was attempted 2-5 min later; CPR was not performed during VF episodes. Countershock was followed by asystole or a pulseless bradyarrhythmia in all animals. Immediate endocardial pacing (0.3-5 mA) produced electrical capture but did not result in arterial pressure pulses in any animal. After pacing, conventional CPR was performed for 2 min or until restoration of spontaneous circulation (ROSC). During CPR, the diastolic coronary perfusion gradient (Ao-RA) was-26-+7 mm Hg (mean-+SD) and CSQ was 0.18-+.08 mL/min/g (42-+23% of control). ROSC followed CPR of less than 2 min duration in 40% of VF study episodes. If ROSC did not follow 2 min of CPR, epinephrine (1 mg), isoproterenol (20 ~g/kg), or glucagon (1 mg) was given IV. Epinephrine increased the CPR Ao-RA gradient to 39 -+ 4 mm Hg and CSQ to 0.51 -+ .09 mL/min/g (118 -+ 16% of control) (all values P < .05 when compared to CPR alone). Isoproterenol decreased the CPR diastolic Ao-RA gradient to 19-+ 10 mm Hg and CSQ to 0.14+.02 mL/min/g. Glucagon initially produced an increase in heart rate, followed by an increase in LV dp/dt and spontaneous arterial pressure fluctuations. Three animals were beta-blocked (IV propranolol 0.5 mg/kg) prior to VF, and postcountershock brady/asystole uniformly responded to glucagon administration, ie, ROSC. Conclusions: 1} effective CPR alone may facilitate resuscitation from postcountershock brady/asystole; 2) immediate pacing alone is of no value in the management of postcountershock brady/asystole; 3) epinephrine improves but isoproterenol decreases myocardial blood flow during resuscitation of brady/asystole. Isoproterenol is of limited value in this setting and may be contraindicated; and 4) glucagon may be of value in the management of postcountershock rhythm disturbances , particularly in the setting of prearrest beta-adrenergic blockade.