Plasma catecholamine levels rise markedly with cardiac arrest and attempted resuscitation. We examined whether epinephrine (EPI) or norepinephrine (NE) plasma concentrations could predict resuscitation outcome. In nine mongrel dogs, EPI and NE levels were drawn before cardiac arrest and after 8 and 14 min of cardiac arrest and CPR. Intravenous EPI (1 mg) was given 1 min before the last plasma level was drawn. Catecholamines were quantitated by high-performance liquid chromatography with triple-electrode coulometric electrochemical detection. Plasma catecholamines increased significantly with cardiac arrest, EPI levels increased from a control level of 15.9 ± 3.0 to 396.0 ± 63.3 pmol/ml after 8 min of cardiac arrest (p < .05), and NE levels similarly increased from 4.4 ± 1.7 to 66.5 ± 12.0 pmol/ml (p < .01). Neither the absolute catecholamine plasma concentration nor the response to cardiac arrest of the endogenous catecholamine concentrations could predict outcome, but catecholamine responses to exogenous EPI did correlate with outcome. Animals which were subsequently resuscitated had a greater increase in the plasma EPI concentrations after exogenous EPI than animals that were not resuscitated, a 53-fold vs. a 23-fold increase (p < .05). Successfully resuscitated animals also had increased NE levels after exogenous EPI, while unsuccessfully resuscitated animals had either no change or a decrease (p < .02). Successfully resuscitated animals had an increase in coronary perfusion pressure (p < .01) in response to exogenous EPI, in contrast to those that were not resuscitated. This suggests that the exogenous administration of EPI during prolonged CPR is beneficial despite markedly elevated endogenous catecholamine levels.