BACKGROUND: Traumatic brain injury (TBI) and subarachnoid hemorrhage (SAH) are the most common causes of severe acute brain injury in younger Intensive Care Unit (ICU) patients. Dysglycemia (abnormal peak glycemia, glycemic variability, mean glycemia, nadir glycemia) is common in these patients but its comparative outcome associations are unclear. METHODS: In a retrospective, cross-sectional, study of adults admitted to Australian and New Zealand ICUs with TBI and SAH from 2005 to 2015, we studied the relationship between multiple aspects of early (first 24 hours) dysglycemia and mortality and compared TBI and SAH patients with the general ICU population and with each other. RESULTS: Among 670,301 patients, 11,812 had TBI and 6,098 had SAH. After adjustment for illness severity, we found that the mortality rate increased with each quintile of glycemia for each aspect of early dysglycemia (peak glycemia, glycemic variability, mean glycemia, nadir glycemia; P<0.0001 for all). This increased risk of death was greater in TBI and SAH patients than in the general ICU population. Moreover, it was stronger for mean glycemia (increase in mortality from 9.2% in the lowest quintile to 15.1% in general ICU patients compared with an increase in mortality from 4.4% to 49.0% for TBI and SAH patients; P<0.0001). Finally, in TBI patients, this relationship was significantly stronger than in SAH patients (P<0.0001). CONCLUSIONS: In TBI and SAH patients, greater dysglycemia is associated with greater mortality. This association is significantly stronger than in the general population and it is significantly stronger in patients with TBI compared with SAH.