Purpose: The study aimed to investigate the association between blood glucose or lactate and theoutcomes of severe traumatic brain injury (TBI), and to evaluate the effect of mild hypothermia therapyon glucose and lactate levels.Methods: Eighty-one patients with TBI were randomly divided into normothermia (n = 41) and mildhypothermia (n = 40) group. Body temperature of hypothermia group was maintained at 32.7°C for72 hours. Arterial blood glucose and lactic acid were determined before and after hypothermia therapy.Glasgow Outcome Scale (GOS) score was assessed 3 months after the treatment.Results: The mean glucose (7.04 ± 0.51 vs 9.71 ± 1.63 mmol/L, P b .05) in the hypothermia group waslower than in the normothermia group after hypothermia therapy. There were more patients with goodneurologic function (GOS 4-5) in the hypothermia group than in the normothermia group (75.0% vs51.2%, P = .038). Multivariate regression analysis showed that blood glucose greater than 10 mmol/L(adjusted risk ratio, 5.7; 95% confidence interval, 1.4-13.2; P b .05) was an independent predictor forpoor neurologic outcomes in these patients, and hypothermia therapy was an independent predictor forfavorable outcomes (risk ratio, 4.9; 95% confidence interval, 1.0-15.6; P b .05). No significantassociation between lactate and GOS scores was identified in the multivariate analysis.Conclusion: Hyperglycemia after TBI was associated with poor clinical outcomes, but the predictivevalue of blood lactate level requires further investigation. Hypothermia therapy improves neurologicoutcomes in patients with severe TBI, and reduction in blood glucose may be partially responsible forthe improved outcomes.