We and others hypothesized that noxious substances released after prolonged cardiac arrest from malfunctioning liver, kidneys, or intestine (e.g. bacterial toxins, aromatic amino acids), might hamper recovery of the brain. The highly detoxifying effect of hemabsorption (i.e. hemoperfusion) with microencapsulated activated carbon has been demonstrated in other diseases. We used our dog model of ventricular fibrillation cardiac arrest of 15 min (n = 2 x 4) or 12.5 min (n = 2 x 6), reversed by brief (high flow) cardiopulmonary bypass (CPB). In half of the dogs in each insult group, a charcoal filter (HemoKart) was inserted into the circuit of CPB at low flow, from start of reperfusion to 4 h. Intermittent positive pressure ventilation was to 20 h and intensive care to 96 h after cardiac arrest. Bacterial blood cultures were positive in most of the dogs in both groups 30 min to 20 h after cardiac arrest (but not later) and were uninfluenced by hemabsorption. In the control groups to 4 h after cardiac arrest, serum levels of potentially injurious aromatic amino acids (e.g. phenylalanine, tyrosine) and of branched-chain/aromatic amino acid ratios, remained unchanged. From 12 to 48 h after cardiac arrest, aromatic amino acid levels increased (worsened). The branched-chain/aromatic amino acid ratios changed accordingly in the opposite direction. In the hemabsorption groups to 4 h after cardiac arrest, all amino acid levels were reduced, aromatic amino acids more so than branched-chain amino acids, thus increasing (improving) the ratio, compared with controls (P < 0.01). There was no group difference after discontinuance of hemabsorption at 4 h. Outcome in terms of overall performance categories and neurologic deficit scores from 24 to 96 h and brain histopathologic damage scores 96 h after cardiac arrest, were not significantly different between groups. The lack of a beneficial outcome effect of hemabsorption to 4 h after cardiac arrest does not support the self-intoxication hypothesis. The amino acid levels later after cardiac arrest suggest that more prolonged hemabsorption and more encompassing detoxification treatments, such as plasma phoresis or total body blood washout, might be evaluated.