Hyperglycemia commonly occurs in patients who are acutely ill, in a variety of clinical situations. Generally, moderate hyperglycemia in critically ill patients was thought to be beneficial; however, the degree of hyperglycemia on admission and the duration of hyperglycemia during critical illness are now recognized markers of adverse outcome. The use of insulin therapy to maintain normoglycemia for at least a few days improves survival and reduces morbidity in patients who are in a surgical intensive care unit (ICU), as shown by a large, randomized, controlled study. These results were recently confirmed by two studies--a randomized, controlled study of patients in a medical ICU, and a prospective, observational study of a heterogeneous patient population admitted to a mixed medical and surgical ICU. Results of multicenter trials that investigated tight blood-glucose control in critically ill patients are, however, still lacking. While we await those multicenter results, the current evidence favors the control of blood glucose levels in the ICU. Indeed, the studies showed that many lives are saved with this intervention, despite an increased incidence of hypoglycemia. Prevention of glucose toxicity by strict glycemic control (but also other metabolic and nonmetabolic effects of insulin) contribute to these clinical benefits.