Kinetic parameters were studied to determine the effectiveness of hemoperfusion in removing ethychlorvynol from the plasma and red blood cells (RBC) of intoxicated dogs. Perfusion columns contained polystyrene/divinyl benzene resin (XAD-4 Amberlite. Column clearances of ethchlorvynol averaged 96.5 +/- 0.4% of the plasma flow rate (mean +/- SEM, 9 dogs). Plasma ethchlorvynol t 1/2's during preperfusion periods averaged 94.1 hr. During hemoperfusion, t 1/2's averaged 3.8 hr, or 90.3 hr shorter than at the endogenous rate of detoxication. There was no significant difference between preperfusion and postperfusion half lives. An estimate based on plasma column clearance suggests that 1.5 +/- 0.1 g ethchlorvynol, or 19.0 +/- 2.8% of the dose, was removed by hemoperfusion. The amount eluted from the resin was 2.9 +/- 0.3 g (37.2 +/- 5.8% of the dose), or about twice the amount apparent from plasma clearance alone. Further, the volume of distribution of ethchlorvynol was 2.3 +/- 0.2 liters/kg, suggesting significant distribution to intracellular and extravascular compartments. The results show that resin hemoperfusion removes a large fraction of ethchlorvynol from intoxicated dogs, and greatly adds to endogenous mechanisms for elimination. Ethchlorvynol was removed from RBC directly, and ultimately from extravascular sites as well.