Recently, several investigators have begun to question the routine use of sodium bicsrbonate in metabolic acidosis, based on a failure to clearly demonstrate the efficacy of alkali therapy, which includes the production of carbon dioxide and variability of the effect on hemodynamic state. We studied the use of sodium bicarbonate in a canine model of hemorrhagic shock to determine its effect on arterial, mixed venous blood gases and hemodynamic states. Nine adult mongrel dogs were anesthetized with pentothal sodium and mechanical ventilation was adjusted to maintained the PaCO2 at 30 to 35mmHg. Ar Swan-Ganz catheter was inserted via a right femoral vein and the right femoral artery was cannulated for continuous pressure monitoring and intermittent blood sampling. 30 minutes after hemorrhagic shock, sodium bicarbonate (1mEq/kg) was administered and 1, 5, 15, 30 and 60 minutes after administration of sodium biearbonate we analyzed the arterial, mixed venous blood gases and measured hemodynamic states. The results were as follows, 1) The arterial carbon dioxide tensions(PaCO2) of 1,5,15,30 and 60 minutes after administration of sodium bicarbonate were 44,42,41,42 and 46mmHg which increased significantly compared to control value, 33mmHg. 2) The mixed venous carbon dioxide tensions(PvCO2) ofr 1, 5, 15, 30 and 60 minutes after administration of sodium bicarbonste were 57, 55, 56, 55 and 55mmHg which also increased significantly compared to control value, 46mmHg. 3) The mean arterial pressures of 1, 5, 15, 30 and 60 minutes after administration of sodium bicarbonate were 61, 60, 64, 68 and 70mmHg which increased significantly compared to control value, 50mmHg, but there were no increasements of cardiac output. It is undesirable to use sodium bicarbonate routinely during hemorrhagic shock because the use of sodium bicarbonate in metabolic acidosis increased arterial and mixed venous carbon dioxide tension and did not show the improvement of hemodynsmic derangement.