Twenty mongrel dogs, weighing between 7.5 and 13.0 kg were used to investigate the percentage limits permissible for hemodilution using a double-helical reservoir heart-lung machine which has a 1,100 ml of priming volume. In both 40 and 50 per cent groups of intentional hemodilution by 30 minute extracorporeal circulation, remarkable anemia was inevitable and recovery was extremely slow, especially in the 50 per cent dilution group. In both 40 and 50 per cent groups of intentional hemodilutions by 30 minute extracorporeal circulation, metabolic acidosis was observed. In 50 per cent group of intentional hemodilution, no improvement of metabolic acidosis was observed even after perfusion. When sodium bicarbonate was administered to 40 per cent hemodilution group, minimum alterations of acid-base balance and of serum electrolytes were observed during and after extracorporeal perfusion. When sodium bicarbonate was administered to 50 per cent hemodilution group, metabolic acidosis was more evident than in 40 per cent hemodilution group accompanied with an increase in serum sodium concentration and a decrease in serum chloride concentration. These data qualify the use of 40 per cent intentional hemodilution using Lactate Ringer's solution or low molecular weight dextran for 30 minute extracorporeal circulation when sodium bicarbonate is administered in adequate amounts.