This study was designed to determine blood flow to the liver during hypercapnia and combined hypercapnia-hypoxia with the portal vein and hepatic artery intact except for placement of an electromagnetic flow probe around these vessels. Twenty mongrel dogs weighing 30-45 kg were anesthetized with pentobarbital and flow probes and occluders were surgically implanted. Ten of these dogs were subjected to hypercapnia alone. During inspiration of 6% CO2 in room air, portal vein flow increased from 588 +/- 73 ml/min to 731 +/- 113 ml/min (p less than .05), while hepatic artery flow did not change significantly from its control mean of 221 +/- 38 ml/min. In the remaining dogs, inhalation of 6% O2 resulted in a reduction of portal blood flow within 30 min from 527 +/- 55 ml/min to 381 +/- 41 ml/min (p less than .01). Again, mean hepatic artery flow did not increase significantly above its control of 273 +/- 43 ml/min. Subsequent inhalation of 6% CO2 plus 6% O2 (combined hypercapniahypoxia) for 30 min in these same animals resulted in a significant increase of portal vein blood flow from 514 +/- 46 ml/min to 716 +/- 116 ml/min (p less than .05). Thus, hypercapnia alone increases total liver blood flow, primarily by an increase in portal vein flow. Hypoxia results in a decrease in portal vein flow. The superimposition of hypercapnia on hypoxia restores blood flow to a level close to that found with hypercapnia alone. Hypercapnia in the range of 63 +/- 4 mmHg PCO2 overwhelms the tendency toward a reduction of portal vein blood flow induced by an arterial PO2 of 42 +/- 5 mmHg in the presence of mild hypocapnia (PCO2 : 30.2 +/- 1 mmHg).