To assess survival and functional outcome in patients endotracheally intubated after ischemic stroke (IS) or spontaneous intracerebral hemorrhage (ICH). Endotracheal intubation is both a necessary life support intervention and a measure of severity in IS or ICH. Knowledge of associated clinical variables may improve the estimation of early prognosis and guide management in these patients. We reviewed 131 charts of patients with IS or ICH who were admitted to the Neurosciences Intensive Care Unit at Duke University Medical Center between July 1994 and June 1997 and required endotracheal intubation. Stroke risk factors, stroke type (IS or ICH) and location (hemispheric, brainstem, or cerebellum), circumstances surrounding intubation, neurologic assessment (Glasgow Coma Score [GCS] and brainstem reflexes), comorbidities, and disposition at discharge were documented. Survivors were interviewed for Barthel Index (BI) scores. Survival was 51% at 30 days and 39% overall. Variables that significantly correlated with 30-day survival in multivariate analysis included GCS at intubation (p = 0.03) and absent pupillary light response (p = 0.008). Increase in the GCS also correlated with improved functional outcome measured by the BI (p = 0.0003). In patients with IS, age and GCS at intubation predicted survival, and in patients with ICH, absent pupillary light response predicted survival. Predictors for mortality differ between patients with IS and ICH; however, decreased level of consciousness is the most important determinant of increased mortality and poor functional outcome. Absent pupillary light responses also correspond with a poor prognosis for survival, but further validation of this finding is needed.