The arteriovenous difference (A-V) method was utilized to assess the permeability of the blood-brain barrier to lactic acid in paralyzed and artificially ventilated newborn dogs. A femoral artery and the sagittal sinus were cannulated to sample arterial and cerebral venous blood simultaneously for measurements of glucose and lactate during normoglycemia, normoglycemia and hyperlactatemia insulin-induced hypoglycemia, or hypoglycemia and hyperlactatemia. During normoglycemia, arterial lactate concentrations remained less than 2 mmoles/liter for up to 2 h; mean A-V lactate was essentially zero. Arterial lactate increased up to 8 mmoles/liter during intravenous infusion of neutralized 10 mM L-lactic acid. During hyperlactatemia, the A-V lactate was directly proportional to the arterial concentration of the metabolite, a finding which is consistent with transport into brain either by simple diffusion or via a carrier with saturability greater than 8 simple diffusion or via a carrier with saturability greater than 8 mmoles/liter. During hypoglycemia (mean arterial glucose=27 mg/dl), A-V glucose was reduced by 71% with a significant increase in A-V lactate at an arterial lactate level of 1.3 mmoles/liter. Hyperlactatemia combined with hypoglycemia resulted in A-V lactate which was 2-3 fold greater than during normoglycemia at similar arterial lactate concentrations. Brain/blood lactate ratios declined by 83% during hypoglycemia compared with normoglycemic ratios, indicating that, once in brain, lactic acid was actively consumed for oxidative processes. These experimental observations may have clinical relevance in newborn human infants when concentrations of lactate in blood often approach or even exceed those of glucose.