Critical illness is characterized by a hypermetabolic state associated with increased mortality, which is partly ascribed to the occurrence of hyperglycemia caused by enhanced endogenous glucose production and insulin resistance (IR). Insulin resistance is well described in patients after surgery and trauma. However, it is less clearly quantified in critically ill medical patients. In this clinical cohort study, IR (M value) was quantified in 40 critically ill medical patients and 25 matched, healthy controls by isoglycemic hyperinsulinemic clamps after an overnight fast on the day after admission to a medical intensive care unit. Energy and substrate metabolism were measured by using indirect calorimetry in the patients before and during the clamp. The severity of illness was assessed by the acute physiology and chronic health evaluation (APACHE) III score. M values of critically ill medical patients were significantly lower compared with healthy controls (2.29 +/- 1.0 and 7.6 +/- 2.9 mg/kg per minute, respectively; P < .001) and were closely related to APACHE III scores (r = -0.43, P < .01), body mass index (r = -0.41, P < .01), and resting energy expenditure (r = 0.40, P < .05). The M value was not associated with age, basal glucose concentrations, and respiratory quotient, and it did not differ among patients with various admission diagnoses. In conclusion, insulin sensitivity was found to be reduced by 70% in critically ill medical patients. The severity of IR was associated with the severity of illness, body mass index, and resting energy expenditure, but not with substrate oxidation rates. In addition, the severity of IR did not vary among patients with different admission diagnoses.