T hough we do not disagree with the need for tight glycemic control in the hospitalized patient, as rightly recommended in the review article and editorial of the supplement, Avoiding Complications in the Hospitalized Patient: The Case for Tight Glycemic Control, 1,2 we feel that any review of hyperglycemia in the hospital setting is incomplete without mention of an important study performed by Van den Berghe and colleagues in a medical ICU. 3 The results of this study are somewhat conflicting with the recent practice of tight control of hyperglycemia in critically ill patients. Intensive insulin therapy reduced morbidity but did not significantly reduce mortality in 1200 medical ICU patients treated in this study as compared with the conventional therapy. Subgroup analysis revealed that patients treated in the medical ICU for more than 3 days did have mortality and morbidity benefit, which was not noticed in the all-patient analysis. Subgroup analysis of patients treated in the medical ICU for fewer than 3 days revealed a rather perplexing result. A nonsignificant increase in mortality was noted in this subgroup. The reasons for such finding are unclear and debatable. Somewhat surprisingly, intensive insulin therapy failed to significantly reduce bacteremia in the medical ICU. This is in contrast with a study done in the surgical ICU, where a significant reduction in bacteremia was noticed with intensive insulin therapy. 4 The results of this study are confusing, and hopefully further clinical trials would clarify the issues. Certainly, the need for tight control of hyperglycemia is proven, but we should acknowledge results of this study as well. In our view, this study is certainly worth mentioning in any comprehensive review of hyperglycemia in critically ill patients.