Abnormal blood glucose levels are common during critical illness and are associated with outcomes that correspond to a J-shaped curve, the lowest risk associated with normoglycemia. Three proof-of-concept randomized-controlled-trials performed in the surgical, medical, and pediatric intensive care units of the Leuven University Hospital in Belgium demonstrated that maintaining strict age-adjusted normal fasting levels of glycemia (80-110 mg/dl in adults, 70-100 mg/dl in children, 50-80 mg/dl in infants) with intensive insulin therapy reduced morbidity and mortality as compared with tolerating stress hyperglycemia as a potentially beneficial response. Recently, concern has risen about the safety of this intervention, as a multicenter adult study reported an, as yet unexplained, increased mortality with targeting normoglycemia as compared with an intermediate blood glucose level of around 140 mg/dl. This apparent contradiction may be explained by several methodological differences among studies, comprising, among others, different glucose target ranges in the control groups, different feeding policies, and variable accuracy of tools used for glucose measurement and insulin infusion. Hence, efficacy and safety of intensive insulin therapy may be affected by patient-related and ICU setting-related variables. Therefore, no single optimal blood glucose target range for ICU patients can be advocated. It appears safe not to embark on targeting "age-normal" levels in intensive care units (ICUs) that are not equipped to accurately and frequently measure blood glucose, and have not acquired extensive experience with intravenous insulin administration using a customized guideline. A simple fallback position could be to control blood glucose levels as close to normal as possible without evoking unacceptable blood glucose fluctuations, hypoglycemia, and hypokalemia.