In expts. on 3 humans with acidosis induced by oral NH4Cl the bicarbonate index of saliva (ratio of bicarbonate to rate of secretion) temporarily decreased with a marked decrease of plasma bicarbonate or showed little correlation with the plasma bicarbonate level. The effect of pH of the blood on salivary bicarbonate is indicated by an expt. on 2 patients, which showed after 35 min. of forced respiration a marked increase in the bicarbonate index of saliva and a decrease of plasma bicarbonate. To investigate further the source of salivary bicarbonate, rabbits weighing 3 kg. were anesthetized with Nembutal and given subcut. injn. of pilocarpine HC1. NaHCl4O3 or carboxyl-labeled lactate was given intraven. Saliva and blood were collected at intervals over 2 hrs. In 10 expts. with bicarbonate the value of the specific activity of salivary bicarbonate was slightly greater than the value for plasma during the first 30 min. Thereafter the value for saliva exceeded the activity for plasma and in all expts. was twice the plasma value at 120 min. It appears that there is no equilibrum between bicarbonate of plasma and saliva and that all of the salivary bicarbonate is not accounted for by plasma. In 2 expts. with lactate the activities due to bicarbonate and other compounds were distinguished by counting samples before and after aeration under acid conditions. The value of the specific activity of salivary bicarbonate remained higher than the value for plasma throughout the expts. Evidence is presented that lactate, or some compund derived from it, is metabolized in the salivary gland to form bicarbonate.