OBJECTIVE: To evaluate the discriminating ability of various specific endocrine studies on patient outcome from the intensive care unit (ICU). DESIGN: Prospective cohort study of patients requiring intensive care. SETTING: Adult medical and coronary care units in a military referral hospital. PATIENTS: A total of 61 consecutive patients requiring intensive care over a 5-month period and 20 control subjects. INTERVENTIONS: Patients were evaluated within 24 hrs of ICU admission (day 1) with determination of the following variables: serum triiodothyronine, thyroxine, triiodothyronine resin uptake, thyrotropin, luteinizing hormone, follicle-stimulating hormone, testosterone, basal cortisol, adrenocorticotropic hormone-stimulated cortisol, cortisol increment, and Acute Physiology and Chronic Health Evaluation (APACHE II) score. A total of 24 hrs later (day 2), the same battery of tests was repeated with the exception of the adrenocorticotropic hormone-stimulated cortisol, cortisol increment, and APACHE II score. Individual variables were compared between survivors and nonsurvivors. MEASUREMENTS AND MAIN RESULTS: The best discriminators of patient outcome in descending order were the basal serum cortisol and triiodothyronine concentrations obtained on day 2 and the APACHE II score with predictive abilities of 81%, 74%, and 70%, respectively. No combination of variables was superior to the day 2 basal cortisol concentration for discrimination of outcome. CONCLUSIONS: The basal cortisol and triiodothyronine concentrations obtained from blood samples collected within 48 hrs of ICU admission appear to be better discriminators of patient outcome than the APACHE II score.