Hyperglycemia upon hospital admission has been associated with poorer neurologic outcomes in patients with brain injury, but this relationship has not been well defined. To evaluate the relationship of hyperglycemia and severity of illness to neurologic outcome, the authors examined Surgical Intensive Care Unit (SICU) records for a 6 month period at a Level I trauma center. Of 276 trauma admissions, 97 patients had intracranial injuries. The peak glucose determination on the first day of admission was correlated with the Glasgow Coma Scale (GCS) score upon admission and discharge from the SICU and with severity of illness as measured by the Simplified Acute Physiology Score (SAPS). The mean admission GCS was 10.6 (+/- 0.49 S.E.M.), the mean glucose on the first SICU day was 146 (+/- 7.7 S.E.M.), and the mean peak glucose was 176 (+/- 8.2 S.E.M.). The peak glucose was inversely related to both GCS on admission and GCS at discharge (P < 0.001). However, stepwise multiple regression analysis revealed that the best single predictor of GCS at discharge was the GCS on admission. The next best predictor was the SAPS on the first SICU day. Peak glucose did not add to the power of admission GCS and SAPS to predict neurologic outcome. Peak glucose levels in brain-injured patients may simply reflect severity of illness and injury that is better represented by SAPS.