Among 700,000 new and recurrent ischemic stroke patients per year, forty percent are hyperglycemic on admission. In-vitro, hyperglycemia is toxic to neurons. Acute ischemic stroke patients who are hyperglycemic on admission experience higher morbidity and mortality. Results of multiple trials have provided no evidence supporting benefit in achieving normoglycemia. On the contrary, there is some evidence that tight glycemic control in acute brain injury is associated with poor outcome. Current consensus derived guidelines from the American Heart Association/American Stroke Association recommend an upper limit of blood glucose of 140–180 mg/dl, as there is no evidence to support strict control. The lack of improved outcomes with normoglycemia in this population dictates reconsideration of assumptions regarding the underlying pathophysiology of hyperglycemia. Review of the current data suggests there are two distinct pathophysiologic entities of hyperglycemia in acute ischemic stroke patients: diabetic and non-diabetic. We propose that the lack of positive results from well-designed intention-to-treat trials in hyperglycemic acute ischemic stroke patients could be attributed to treating these distinct groups as one.